COMPOUNDS AND METHODS FOR SYNERGISTIC ACTIVATION OF M CHANNELS
A method of activating heteromeric KCNQ2/3, KCNQ3/5, or KCNQ4/5 voltage-gated potassium channels in a cell membrane is described, as are compositions for use with same, as well as methods of reducing neuronal excitability and of ameliorating symptoms of epilepsy, anxiety, neuropathic pain, hypertension, cardiovascular disease, a neurodegenerative disorder, alcohol withdrawal, cancer, inflammation, or ophthalmic disease in a subject. These KCNQ2/3 (or KCNQ3/5 or KCNQ4/5) channels are heteromers that comprise KCNQ2 (or KCNQ5) subunits and KCNQ3 (or KCNQ4) subunits. In one embodiment, the method comprises: (a) contacting the cell membrane with a first agent that binds a KCNQ2 (or KCNQ5) subunit; and (b) contacting the cell membrane with a second agent that binds a KCNQ3 (or KCNQ4) subunit.
Latest THE REGENTS OF THE UNIVERSITY OF CALIFORNIA Patents:
This invention was made with Government support under GM115189 awarded by the National Institutes of Health and National Institute of General Medical Sciences. The Government has certain rights in the invention.
BACKGROUNDKCNQ channels are expressed in many different tissues and serve crucial roles. Opening KCNQ channels with small molecules has therapeutic potential, and KCNQ openers, e.g., retigabine, have been previously approved clinically for epilepsy. However, retigabine was recently taken off the market because of unexpected side effects thought to be unrelated to its targeting of KCNQ channels.
There remains a need for safe, effective, and reliable KCNQ openers, and methods for managing neuronal excitability as well as electrical activity of other cells types outside the brain, including but not limited to cardiac myocytes, epithelial and endothelial cells.
SUMMARYDescribed herein is a method of activating heteromeric KCNQ2/3, KCNQ3/5, or KCNQ4/5 voltage-gated potassium channels in a cell membrane. These KCNQ2/3 (or KCNQ3/5 or KCNQ4/5) channels are heteromers that comprise KCNQ2 (or KCNQ5) subunits and KCNQ3 (or KCNQ4) subunits. In one embodiment, the method comprises: (a) contacting the cell membrane with a first agent that binds a KCNQ2 (or KCNQ5) subunit; and (b) contacting the cell membrane with a second agent that binds a KCNQ3 (or KCNQ4) subunit. Representative embodiments of the first agent that binds a KCNQ2 (or KCNQ5) subunit include, but are not limited to, one or more agents selected from mallotoxin (MTX), isovaleric acid (IVA), and zinc pyrithione. Representative embodiments of the second agent that binds a KCNQ3 subunit include, but are not limited to, one or more agents selected from retigabine (RTG) and derivatives thereof, gabapentin, β-hydroxybutyric acid (BHB), and γ-Amino-p-hydroxybutyric acid (GABOB).
Also provided is a composition that activates KCNQ2/3, KCNQ3/5, or KCNQ4/5 voltage-gated potassium channels in a cell membrane, said channels comprising KCNQ2 subunits, KCNQ3 subunits, KCNQ4 subunits, and/or KCNQ5 subunits. In one embodiment, the composition comprises a first agent that binds a KCNQ2 subunit or a KCNQ5 subunit, and a second agent that binds a KCNQ3 subunit or a KCNQ4 subunit, wherein the first and second agents comprise non-identical agents. In some embodiments, the first agent that binds a KCNQ2 or KCNQ5 subunit is one or more agents selected from mallotoxin (MTX), isovaleric acid (IVA), zinc pyrithione (ZnPy), 2-fluorophenylglycine (2FPG), aloperine, 4-(fluorophenyl)glycine (4FPG), ICA-069673, ICA-27243, ztz-240, and (E)-2-dodecenal (E2D); retigabine (RTG) and derivatives thereof, gabapentin (GBP), β-hydroxybutyric acid (BHB), γ-Amino-p-hydroxybutyric acid (GABOB). In some embodiments, the second agent that binds a KCNQ3 or KCNQ4 subunit is one or more agents selected from retigabine (RTG) and derivatives thereof, gabapentin (GBP), β-hydroxybutyric acid (BHB), γ-Amino-β-hydroxybutyric acid (GABOB), and N-(fluorophenyl)-N-(methylsulfonyl)glycine (3FMSG). In some embodiments, the first agent is MTX and/or IVA, and the second agent is RTG. In other embodiments, the first agent is 2FPG, and the second agent is GBP or 3FMSG.
In some embodiments, the first and second agents are present in an amount ranging from 1 to 100 μM. In other embodiments, the agents are present in an amount of 10 μM or less. The composition can be formulated for administration in accordance with the methods described herein. For example, the amount of agents present in the composition can be determined for optimal delivery of the intended daily dosage. In one example, the daily dosage is 10 to 1000 mg. In another example, the daily dosage is 20 to 240 mg.
Also provided is a method of activating KCNQ2/3, KCNQ3/5, or KCNQ4/5 voltage-gated potassium channels in a cell membrane, said channels comprising KCNQ2 or KCNQ5 subunits, and further comprising KCNQ3 or KCNQ4 subunits, the method comprising contacting the cell membrane with a composition as described herein. In some embodiments, the cell is a neuron, myocyte, epithelial cell, or endothelial cell. In one particular embodiment, the cell is a neuron. In some embodiments, the cell membrane potential is −80 mV to +40 mV during the contacting of the cell membrane with the composition. In other embodiments, the cell membrane potential is −80 mV to −40 mV.
Additionally provided is a method of reducing neuronal excitability in a subject, the method comprising administering to the subject a composition as described herein. Another method provided is a method of ameliorating symptoms of epilepsy, anxiety, neuropathic pain, hypertension, cardiovascular disease, a neurodegenerative disorder, alcohol withdrawal, cancer, inflammation, or ophthalmic disease in a subject, the method comprising administering to the subject a composition as described herein. In some embodiments of these methods, the first and second agents of the composition are each administered at a dose of 10-1000 mg/day. In other embodiments, the first and second agents are each administered at a dose of 10-500 mg/day. In some embodiments, the dose is 20-240 mg/day.
In a representative example, the channel is a KCNQ2/3 channel, and the combination of agents comprises one or more agents that bind KCNQ2 and one or more agents that bind KCNQ3. In one such example, the agent that binds a KCNQ3 subunit is RTG, and the RTG is administered at a dose of 20-240 mg/day. In one embodiment, the second agent that binds a KCNQ3 subunit is RTG, and the RTG is administered at a dose of 20-120 mg/day. In one example, the agent that binds a KCNQ3 subunit is gabapentin.
In another representative example, the channel is a KCNQ3/5 channel, and the combination of agents comprises one or more agents that bind KCNQ3 and one or more agents that bind KCNQ5. In one such example, the agent that binds KCNQ3 is RTG, and the agent that binds KCNQ5 is aloperine. In yet another representative example, the channel is a KCNQ4/5 channel, and the combination of agents comprises one or more agents that bind KCNQ4 and one or more agents that bind KCNQ5.
In some embodiments, the contacting steps occur simultaneously. In other embodiments, the contacting occurs near-simultaneously, or sequentially. In a typical embodiment, the cell membrane potential is below +40 mV during the contacting steps. The cell membrane potential can be negative, such as −40 mV or −80 mV, for example.
In some embodiments of the methods for reducing neuronal excitability and for ameliorating symptoms of one of the aforementioned diseases or disorders, the heteromeric channel is KCNQ3/KCNQ5, and the combination of activators comprises one or more compounds that preferentially activate KCNQ3 (e.g., RTG) and one or more compounds that preferentially activate KCNQ5 (e.g., MTX) within KCNQ3/5 heteromers.
The invention thus provides kits and compositions for use in carrying out the methods described herein. In some embodiments, such kits and compositions comprise a first agent that binds a KCNQ2 subunit of a KCNQ2/3 voltage-gated potassium channel and a second agent that binds a KCNQ3 subunit of a KCNQ2/3 voltage-gated potassium channel. In some embodiments, the kits and compositions comprise a first agent that binds a KCNQ3 subunit of a KCNQ3/5 voltage-gated potassium channel and a second agent that binds a KCNQ5 subunit of a KCNQ3/5 voltage-gated potassium channel. In some embodiments, the kits and compositions comprise a first agent that binds a KCNQ4 subunit of a KCNQ4/5 voltage-gated potassium channel and a second agent that binds a KCNQ5 subunit of a KCNQ4/5 voltage-gated potassium channel. The first and second agents can be provided in a single composition, or, in some embodiments, provided separately. For example, a kit may comprise a plurality of containers, each comprising one or more agents that bind a heteromeric KCNQ voltage-gated potassium channel.
The invention is based on the discovery that surprising advantages can be achieved by activating heteromeric KCNQ2/3, KCNQ3/5, or KCNQ4/5 voltage-gated potassium channels in a cell membrane. The use of combinations of first and second agents that bind both the KCNQ2 and KCNQ3 subunits of the channels provides an unexpectedly synergistic effect. In some cases, triple combinations provide even greater synergy, e.g., two KCNQ2 openers (IVA+MTX) in combination with a KCNQ3 opener (such as RTG), results in a stable open state that locks the KCNQ2/3 channel open in a voltage-independent manner. The synergistic activation of the different isoforms in the KCNQ heteromer produced by double or triple compound combinations increases the potency and efficacy of treatments designed to target M channels, and allows for treatment using lower doses of KCNQ openers that have unwanted side effects and/or toxicity.
DefinitionsAll scientific and technical terms used in this application have meanings commonly used in the art unless otherwise specified. As used in this application, the following words or phrases have the meanings specified.
As used herein, “activating” an M channel, such as a KCNQ2/3 voltage-gated potassium channel, means opening the channel sufficiently that current passes through the channel.
As used herein, “a” or “an” means at least one, unless clearly indicated otherwise.
As used herein, to “prevent” or “protect against” a condition or disease means to hinder, reduce or delay the onset or progression of the condition or disease.
As used herein, “treating” a disease or condition means to ameliorate symptoms and/or delay progression of the disease or condition.
Kits and CompositionsThe invention provides kits and compositions, which comprise a set or mixture of agents as described herein, and optionally, one or more suitable containers housing agents of the invention. Kits and compositions of the invention optionally further comprise a pharmaceutically acceptable carrier or excipient. The kit or composition can optionally include a buffer.
Also provided is a pharmaceutical composition comprising a therapeutically effective amount of one or more agents as described herein, and, optionally, a pharmaceutically acceptable carrier. In one embodiment, the invention provides an injectable, stable, sterile composition comprising an active compound(s), or a salt thereof, in a unit dosage form in a sealed container. The compound or salt thereof can be provided in the form of a lyophilizate which is capable of being reconstituted with a suitable pharmaceutically acceptable carrier to form a liquid composition suitable for administration to a subject. The unit dosage form may in one variation comprise from about 10 mg to about 10 grams of the compound or salt thereof. In some embodiments the dosage comprises from about 1 to about 100 mg of the compound or salt thereof. When the compound or salt thereof is substantially water-insoluble, a sufficient amount of emulsifying agent which is physiologically acceptable may be employed in sufficient quantity to emulsify the compound or salt in an aqueous carrier.
In some embodiments, the agents of the kit or composition comprise specific combinations as described herein. The kits or compositions can, in some representative embodiments, include one or more agents selected from mallotoxin (MTX), isovaleric acid (IVA), zinc pyrithione (ZnPy), 2-fluorophenylglycine (2FPG), aloperine, 4-(fluorophenyl)glycine (4FPG), ICA-069673, ICA-27243, ztz-240, and (E)-2-dodecenal (E2D); retigabine (RTG) and derivatives thereof, gabapentin (GBP), β-hydroxybutyric acid (BHB), γ-Amino-β-hydroxybutyric acid (GABOB), and/or N-(fluorophenyl)-N-(methylsulfonyl)glycine (3FMSG).
MethodsIn some embodiments, the invention provides a method of activating KCNQ2/3, KCNQ3/5, or KCNQ4/5 voltage-gated potassium channels in a cell membrane, said channels comprising KCNQ2 or KCNQ5 subunits, and further comprising KCNQ3 or KCNQ4 subunits. In some embodiments, the method comprises contacting the cell membrane with a composition as described herein. In some embodiments, the cell is a neuron, myocyte, epithelial cell, or endothelial cell. In one particular embodiment, the cell is a neuron. In some embodiments, the cell membrane potential is −120 mV to +40 mV during the contacting of the cell membrane with the composition. In other embodiments, the cell membrane potential is −80 mV to +40 mV, or, in some embodiments, −80 mV to −40 mV.
Additionally provided is a method of reducing neuronal excitability in a subject, the method comprising administering to the subject a composition as described herein. Another method provided is a method of ameliorating symptoms of epilepsy, anxiety, neuropathic pain, hypertension, cardiovascular disease, a neurodegenerative disorder, alcohol withdrawal, cancer, inflammation, or ophthalmic disease in a subject. In a typical embodiment, the method comprises administering to the subject a composition as described herein. In some embodiments of these methods, the first and second agents of the composition are each administered at a dose of 10-1000 mg/day. In other embodiments, the first and second agents are each administered at a dose of 10-500 mg/day. In some embodiments, the dose is 20-240 mg/day. Other exemplary doses are 10-400 mg/day, 10-300 mg/day, 10-200 mg/day, and 10-100 mg/day.
In a representative example, the channel is a KCNQ2/3 channel, and the combination of agents comprises one or more agents that bind KCNQ2 and one or more agents that bind KCNQ3. In one such example, the agent that binds a KCNQ3 subunit is RTG, and the RTG is administered at a dose of 20-240 mg/day. In one embodiment, the second agent that binds a KCNQ3 subunit is RTG, and the RTG is administered at a dose of 20-120 mg/day. In one example, the agent that binds a KCNQ3 subunit is gabapentin.
In another representative example, the channel is a KCNQ3/5 channel, and the combination of agents comprises one or more agents that bind KCNQ3 and one or more agents that bind KCNQ5. In one such example, the agent that binds KCNQ3 is RTG, and the agent that binds KCNQ5 is aloperine. In yet another representative example, the channel is a KCNQ4/5 channel, and the combination of agents comprises one or more agents that bind KCNQ4 and one or more agents that bind KCNQ5.
In some embodiments, the contacting steps occur simultaneously. In other embodiments, the contacting occurs near-simultaneously, or sequentially. In a typical embodiment, the cell membrane potential is below +40 mV during the contacting steps. The cell membrane potential can be negative, such as −40 mV or −80 mV, for example.
In some embodiments of the methods for reducing neuronal excitability and for ameliorating symptoms of one of the aforementioned diseases or disorders, the heteromeric channel is KCNQ3/KCNQ5, and the combination of activators comprises one or more compounds that preferentially activate KCNQ3 (e.g., RTG) and one or more compounds that preferentially activate KCNQ5 (e.g., MTX) within KCNQ3/5 heteromers.
TreatmentThe invention provides methods for administering the agents described herein to a subject, and methods of treating subjects in need thereof. Treatment can be administered in a single dose or as a series of doses administered over time, such as, in one example, daily. Dosage and treatment regimens can be determined by the treating physician, taking into account disease severity, patient condition, and other factors. In one representative embodiment, the agent is RTG, and is administered at a dose selected to avoid unwanted side effects. In a representative example, the dose of RTG is 20 mg/day to 120 mg/day.
For use in the methods described herein, the subject includes any human or non-human animal. The term “non-human animal” includes all vertebrates, e.g., mammals and non-mammals, such as non-human primates, horses, sheep, dogs, cows, pigs, chickens, and other veterinary subjects. In a typical embodiment, the subject is human.
EXEMPLARY EMBODIMENTSEmbodiment 1: A method of activating KCNQ2/3 voltage-gated potassium channels in a cell membrane, said channels comprising KCNQ2 subunits and KCNQ3 subunits, the method comprising: (a) contacting the cell membrane with a first agent that binds a KCNQ2 subunit; and (b) contacting the cell membrane with a second agent that binds a KCNQ3 subunit; wherein the first agent that binds a KCNQ2 subunit is one or more agents selected from mallotoxin (MTX), isovaleric acid (IVA), and zinc pyrithione; and wherein the second agent that binds a KCNQ3 subunit is one or more agents selected from retigabine (RTG) and derivatives thereof, gabapentin, β-hydroxybutyric acid (BHB), and γ-Amino-p-hydroxybutyric acid (GABOB).
Embodiment 2: The method of embodiment 1, wherein the cell is a neuron.
Embodiment 3: The method of embodiment 1, wherein the contacting of steps (a) and (b) occurs simultaneously.
Embodiment 4: The method of embodiment 1, wherein the cell membrane potential is below +40 mV during the contacting of steps (a) and (b).
Embodiment 5: A method of reducing neuronal excitability in a subject, the method comprising administering to the subject a combination of a first agent that binds a KCNQ2 subunit of a KCNQ2/3 voltage-gated potassium channel and a second agent that binds a KCNQ3 subunit of a KCNQ2/3 voltage-gated potassium channel.
Embodiment 6: A method of ameliorating symptoms of epilepsy, anxiety, neuropathic pain, a neurodegenerative disorder, alcohol withdrawal, cancer, inflammation, or ophthalmic disease in a subject, the method comprising administering to the subject a combination of a first agent that binds a KCNQ2 subunit of a KCNQ2/3 voltage-gated potassium channel and a second agent that binds a KCNQ3 subunit of a KCNQ2/3 voltage-gated potassium channel.
Embodiment 7: The method of embodiment 5 or 6, wherein the first agent that binds a KCNQ2 subunit is mallotoxin (MTX), isovaleric acid (IVA), zinc pyrithione, or a combination thereof.
Embodiment 8: The method of embodiment 5 or 6, wherein the second agent that binds a KCNQ3 subunit is one or more agents selected from retigabine (RTG) and derivatives thereof, gabapentin, β-hydroxybutyric acid (BHB), and γ-Amino-β-hydroxybutyric acid (GABOB).
Embodiment 9: The method of embodiment 8, wherein the second agent that binds a KCNQ3 subunit is RTG, and wherein the RTG is administered at a dose of 20-120 mg/day.
Embodiment 10: The method of embodiment 8, wherein the agent that binds a KCNQ3 subunit is gabapentin.
Embodiment 11: The method of any of the preceding embodiments, wherein the heteromeric channel is KCNQ3/KCNQ5, and the combination of agents comprises one or more agents that bind KCNQ3 and one or more agents that bind KCNQ5.
Embodiment 12: The method of embodiment 11, wherein the agent that binds KCNQ3 is RTG, and the agent that binds KCNQ5 is MTX.
EXAMPLESThe following examples are presented to illustrate the present invention and to assist one of ordinary skill in making and using the same. The examples are not intended in any way to otherwise limit the scope of the invention. Supplementary figures and tables noted in and confirming the data described the following Examples can be accessed via the online versions of the corresponding publications.
Example 1: Anticonvulsants Act Synergistically in a KCNQ Potassium Channel Binding PocketEpilepsy has been treated for centuries with herbal remedies, including leaves of the African shrub Mallotus oppositifolius, yet the underlying molecular mechanisms have remained unclear. Voltage-gated potassium channel isoforms KCNQ2-5, predominantly KCNQ2/3 heteromers, underlie the neuronal M-current, which suppresses neuronal excitability, protecting against seizures. In this Example, in silico docking, mutagenesis and cellular electrophysiology reveal that two components of M. oppositifolius leaf extract, mallotoxin (MTX) and isovaleric acid (IVA), act synergistically to open neuronal KCNQs, including KCNQ2/3 channels. Correspondingly, MTX and IVA combine to suppress pentylene tetrazole-induced tonic seizures in mice, whereas individually they are ineffective. Co-administering MTX and IVA with the modern, synthetic anticonvulsant retigabine creates a further synergy that voltage independently locks KCNQ2/3 open. Leveraging this synergy, which harnesses ancient and modem medicines to exploit differential KCNQ isoform preferences, presents an approach to developing safe yet effective anticonvulsants.
Epilepsy affects over 50 million people worldwide, with approximately 50% being inadequately treated with currently available anti-epileptic drugs1. In the developing world, an estimated 80% of epilepsy patients use herbal remedies for primary healthcare2,3. Extract taken from leaves of the shrub Mallotus oppositifolius has been used traditionally in folk medicine to treat disorders such as seizures in African countries4,5 including Ghana, where it is known as nyanyaforowa (pimpim), and Nigeria, where it is referred to as okpo-biriba. Accordingly, M. oppositifolius extract has been shown to delay the onset, frequency, and duration of seizures in the acute chemoconvulsant (pentylene tetrazole) mouse model5. Despite the clear therapeutic effects of M. oppositifolius, the active anticonvulsant components of this extract have remained unclear, and their molecular targets unknown. Mallotoxin (MTX; a.k.a. rottlerin) is one suggested anti-seizure component of M. oppositifolius. However, a plausible molecular target for MTX that would quell seizures has not previously been identified. In addition, some other plants heavily used in folk medicine also contain MTX but are not traditionally used to treat epilepsy. For example, Mallotus philippensis, a perennial shrub distributed in outer Himalayan lowlands, is reported to possess antifilarial, antifertility, antibiotic, anti-inflammatory, and a range of other properties but is not reportedly used to treat seizures6. We therefore hypothesized that MTX either is not the active anticonvulsant in M. oppositifolius or does not act alone.
Members of the KCNQ (Kv7) subfamily of voltage-gated potassium (Kv) channels are essential for control of cellular excitability and repolarization in a wide range of cell types. Kv channels, including the KCNQs, are composed of tetramers of a subunits each containing six transmembrane segments (S1-S6), split into a voltage-sensing domain (S1-S4) and a pore module (S5 and S6) (
RTG was approved by the FDA in 2011 and was in clinical use as an add-on therapy for the treatment of partial seizures in adults with epilepsy until 2017, when it was withdrawn from the market because of side effects including blue skin discoloration and retinal pigment changes. More recently, the skin discoloration has been found to be reversible after drug discontinuation19,20. RTG activates all neuronally expressed KCNQ isoforms (KCNQ2-5), with a preference for KCNQ3 (ref. 21), and in addition to epilepsy showed promise in treating disorders including anxiety, neuropathic pain, neurodegenerative disorders, cancer, inflammation, and ophthalmic diseases22,23,24,25. New drugs are therefore needed that share mechanistic commonalties with RTG but lack the side effects, and ideally possessing improved efficacy and/or potency, to reduce the required dosage.
This Example shows that two components of M. oppositifolius leaf extract, MTX and IVA, act synergistically in similar binding pockets to activate KCNQ2/3 channels and reduce tonic seizure incidence and related mortality in mice. We also demonstrate that when co-administered, MTX, IVA, and RTG voltage-independently lock KCNQ2/3 open at all voltages. We explain the molecular mechanisms underlying these synergies, which suggest a pathway for developing safer, more effective anticonvulsants.
Results
Multiple M. oppositifolius Leaf Compounds Activate KCNQ2/3
Igwe et al.26 recently identified by mass spectrometry nine primary components of an ethanolic extract of the M. oppositifolius leaf, in addition to MTX, which was previously identified in M. oppositifolius bark and leaves27 (
MTX Potently Activates KCNQ2/3 Channels
MTX, a polyphenol, was the most potent KCNQ2/3 activator in our initial screen (
MTX Preferentially Activates KCNQ2 Channels
With respect to homomeric M-channels, KCNQ2 exhibited the highest MTX sensitivity, with an EC50 of 6.4 μM at −60 mV. KCNQ3* (an expression-optimized KCNQ3-A315T mutant that ensures robust currents)31 was twofold less sensitive than KCNQ2 (EC50, 13.0 μM), while KCNQ4 and KCNQ5 had MTX EC50 values of 20.2 and 67.1 μM, respectively (
MTX Binds Close to the Channel Pore to Activate KCNQ2/3
RTG, an established KCNQ2/3 channel activator, requires KCNQ2-W236, which is located on transmembrane segment 5 (S5;
In silico docking simulations predicted binding of MTX in the region of KCNQ3-W265, but not L314 (
Due to the bulky nature of MTX, we predicted that its binding within a pocket close to S5 would influence pore conformation, potentially altering relative ion permeabilities. To assess this, we first conducted pseudo-bi-ionic substitution experiments, which showed that MTX increased relative permeability of KCNQ2/3 to Na+ and Cs+, and decreased permeability to Rb+, compared to K+ (
In further support of this hypothesis, we found that GABOB (γ-amino-β-hydroxybutyric acid), a high-affinity KCNQ2/3 partial agonist that binds close to KCNQ3-W265 (ref. 28), only partially diminishes the effects of MTX on KCNQ2/3 channels (
M. oppositifolius Component IVA Potently Activates KCNQ2
In addition to MTX, previous gas chromatography-mass spectrometry (GCMS) analysis of M. oppositifolius extract identified nine additional compounds26, three of which we found to activate KCNQ2/3 channels in our initial screen (
As we also observed for MTX, IVA activation was voltage-dependent and had the greatest fold-effect on current at −60 mV (
We also performed dose responses for the remaining KCNQ2/3-active compounds in M. oppositifolius leaf extract, i.e., glutaconic acid, 2-mercaptophenol, and palmitic acid, and in addition tested oleamide because at 100 μM it slightly increased KCNQ2/3 currents at higher voltages (
MTX and IVA Synergistically Activate KCNQ2/3 Channels
Because our data identified MTX and IVA as the most active M. oppositifolius components with respect to KCNQ2/3 activation, and also suggested different binding positions for MTX and IVA (
MTX and IVA Synergistically Protect Against Seizures
The cellular electrophysiology data therefore predicted that MTX and IVA would in combination be necessary and sufficient to confer anticonvulsant activity, if KCNQ2/3 activation was the molecular basis for this therapeutic action of the leaf extract. In mouse pentylene tetrazole (PTZ) chemoconvulsant assays, MTX (20 mg/kg) halved the clonic seizure incidence whereas IVA (20 mg/kg) had no effect. At 10 mg/kg neither compound reduced clonic seizures alone, but halved seizure incidence in combination (
MTX and IVA Combine with RTG to Lock KCNQ2/3 Open
Our results indicate that MTX binds close to the channel pore and senses the KCNQ2/3 S5 tryptophans, while IVA absolutely requires them for binding, and that KCNQ2 is the most MTX- and IVA-sensitive isoform (
Kv channel openers are generally more effective at negative membrane potentials because the lower open probability provides more capacity for augmentation before the maximum open probability is reached (in contrast to positive voltages). However, at extremely hyperpolarized membrane potentials, the capacity of channel openers such as RTG to activate diminishes again (creating a bell-shaped voltage dependence to activation). In the case of KCNQ2/3 this may be because of an inability to open the channel from more stable closed conformations. However, addition of MTX or IVA, and in particular both, to RTG overcame this, resulting in potent current augmentation even at −120 mV. Thus, MTX+IVA+RTG increased KCNQ2/3 current by 60-80-fold at −80 to −120 mV (
MTX and IVA have been tolerated as part of herbal medicine for centuries27,34,35. RTG was in clinical use for 6 years before being withdrawn because of adverse off-target effects, although these are now known to subside following RTG discontinuation19,20. Given the synergy between MTX, IVA, and RTG, we tested whether combining the three at low concentrations could achieve efficacy at potentially tolerable RTG doses. At 1 μM, RTG had negligible effects on KCNQ2/3 (
We further tested whether the heteromeric composition of KCNQ2/3 channels afforded greater sensitivity to the synergistic effects of IVA+MTX+RTG than for the homomers. Studies of relative ion permeabilities for heteromeric versus homomeric channels revealed greater increases in relative Na+ and Cs+ permeability (
We therefore next tested whether homomeric KCNQ channels were as comprehensively activated by MTX+IVA+RTG as was KCNQ2/3. KCNQ3* was the most sensitive of the homomers, and the slowest deactivating at −120 mV of all the homomers in response to MTX+IVA+RTG. Interestingly, even homomeric KCNQ4 and KCNQ5 were activated by the highest triple-drug dose (
Thus, the MTX+IVA+RTG combination leverages the heteromeric composition of KCNQ2/3 channels to exert optimal synergistic effects on channel opening. Data from
We have discovered that IVA and MTX, two components of the traditional African anticonvulsant M. oppositifolius leaf extract, synergistically activate KCNQ2/3 and protect against tonic seizures and associated mortality. MTX, the principal component of phenolic extracts of Mallotus, has other reported biological activities;30,37,38,39 none of them readily explain anticonvulsant efficacy, but may contribute to efficacy in other therapeutic uses of M. oppositifolius, which include treatment of pain, infection, and inflammation40. Historical medicinal usage of Mallotus plants spans West Africa (M. oppositifolius) and also parts of Asia, including Bangladesh, China, and India (M. repandus, M. philippinensis, and others)6, 41. Oral bioavailability of MTX in rats fed Mallotus philippensis extract was previously quantified at >20%, and plasma concentrations exceeded 2 μg/ml, reflecting also the high concentration of MTX in Mallotus; e.g., the MTX content of powder prepared from Mallotus philippensis fruit was quantified as 21.25% w/w42. A plasma concentration of 2 μg/ml is equivalent to 4 μM MTX, a concentration at which we observe KCNQ2/3 activation even by MTX alone.
IVA has been quantified to be ˜12% of the methanol extract of the Mallotus leaf extract26. IVA is also a component of valerian root, an herbal medicine used since ancient Greek and Roman times to treat insomnia, and since medieval times in Europe specifically to treat seizures34,35. Valerian root extract is still used extensively today for anxiety and insomnia, although randomized controlled trials evaluating its efficacy have achieved mixed results43,44,45. It has been estimated that 10 g of valerian root might yield as much as 100 mg of IVA, and that valerian root doses of 30-50 g per day would have the potential for anticonvulsant activity46. Pharmacokinetic studies in humans have been performed for NPS 1776, or isovaleramide, the amide derivative of IVA. Isovaleramide readily passes though biological membranes, and is well tolerated in humans up to at least 2400 mg per day. Absorption is rapid (mean Tmax of 30-45 min) and mean elimination half-life is 2.5 h47. Furthermore, volatile fatty acids similar to IVA, e.g., acetate, readily cross the blood brain barrier48. Taken together, these studies together with our current findings suggest that herbal extracts contain sufficient bioavailable IVA to exert effects on KCNQ2/3 channels, particularly if synergizing with MTX.
Mutagenesis results indicating differential requirement of W236/W265 (
The data suggest a model in which MTX and IVA are able to bind to KCNQ2/3 across a wide range of membrane potentials and induce a stable open conformation. One likely configuration, and potentially the dominant configuration given the aforementioned subunit-specific potencies, is that in KCNQ2/3 heteromers MTX and IVA bind to KCNQ2, and RTG to KCNQ3 (
The molecular strategy we present, utilizing a combination of small molecules from ancient and modem therapeutics, may provide a route to safer, more effective anticonvulsants. Through dual synergies it lowers the effective doses required to achieve similar KCNQ2/3 opening, increases the maximal effects, and also considerably broadens the voltage range across which the drugs are effective (
Channel Subunit cRNA Preparation and Oocyte Injection
cRNA transcripts encoding human KCNA1, KCNQ1, KCNQ2, KCNQ3, KCNQ4, or KCNQ5 were generated by in vitro transcription using the T7 polymerase message machine kit (Thermo Fisher Scientific), after vector linearization, from cDNA sub-cloned into plasmids incorporating X. laevis β-globin 5′ and 3′ UTRs flanking the coding region to enhance translation and cRNA stability. cRNA was quantified by spectrophotometry. Mutant cDNAs were generated by site-directed mutagenesis using a QuikChange kit according to manufacturer's protocol (Stratagene, San Diego, Calif.) and corresponding cRNAs prepared as above. Defolliculated stage V and VI X. laevis oocytes (Ecocyte Bioscience, Austin, Tex.) were injected with Kv channel a subunit cRNAs (10 ng total per oocyte). Oocytes were incubated at 16° C. in Barth's saline solution (Ecocyte) containing penicillin and streptomycin, with daily washing, for 3-5 days prior to two-electrode voltage-clamp (TEVC) recording.
Two-Electrode Voltage-Clamp
TEVC recording was performed at room temperature with an OC-725C amplifier (Warner Instruments, Hamden, Conn.) and pClamp10.2 software (Molecular Devices, Sunnyvale, Calif.) 3-5 days after cRNA injection as described in the section above. Oocytes were placed in a small-volume oocyte bath (Warner) and viewed with a dissection microscope. Chemicals were sourced from Sigma. Bath solution was (in mM): 96 NaCl, 4 KCl, 1 MgCl2, 1 CaCl2. 10 HEPES (pH 7.6). Isovaleric acid, 2-mercaptophenol, 1-Heptene, and 3-ethyl-2-hydroxy-2-cyclopenten-1-one were stored at 4° C. as 5 mM stocks in Ringer's solution. MTX (DMSO), sorbic acid (ethanol), and glutaconic acid (molecular grade H2O) were stored at −20° C. as 1 M stocks. Oleamide was stored as a 1 mM stock in ethanol at 4° C. Palmitic acid was conjugated with bovine serum albumin as a 1 mM stock and stored at −20° C. All compounds were diluted to working concentrations each experimental day. All compounds were introduced to the recording bath via gravity perfusion at a constant flow of 1 ml/min for 3 min prior to recording. Pipettes were of 1-2 MΩ resistance when filled with 3 M KCl. Currents were recorded in response to pulses between −80 and +40 mV at 20 mV intervals, or a single pulse to +40 mV, from a holding potential of −80 mV, to yield current-voltage relationships, current magnitude, and for quantifying activation rate. TEVC data analysis was performed with Clampfit10.2 (Molecular Devices) and Graphpad Prism software (GraphPad, San Diego, Calif., USA); values are stated as mean±SEM. Normalized tail currents were plotted versus prepulse voltage and fitted with a single Boltzmann function:
where g is the normalized tail conductance, A1 is the initial value at −∞, A2 is the final value at +∞, V1/2 is the half-maximal voltage of activation, and Vs the slope factor. Activation and deactivation kinetics were fitted with single-exponential functions.
For relative permeability studies, currents were recorded in response to a single pulse to +40 mV for 5 s, followed by pulses between −80 mV and +40 mV at 20 mV intervals, from a holding potential of −80 mV, to yield a current-voltage relationship. According to the Goldman-Hodgkin-Katz (GHK) voltage equation:
where Erev is the absolute reversal potential and P is the permeability. This permits calculation of the relative permeability of each ion if concentrations on either side of the membrane are known. A modified version of this equation was used here to determine relative permeability of two ions in a system in which only the extracellular ion concentration was known. Thus, relative permeability of Rb+, Cs+, and Na+ compared to K+ ions was calculated for all channels by plotting the //V relationships for each channel with each extracellular ion (100 mM) and comparing them to that with 100 mM extracellular K+ ion to yield a change in reversal potential (ΔErev) for each ion compared to that of K+. Permeability ratios for each ion compared to K+ were then calculated as
Values were compared between channel types and statistical significance assessed using ANOVA.
For calculating Rb+/K+ permeability (
Deactivating currents were fitted to a single-exponential standard function defined as follows:
∫(t)=Σi=1nAie−t/τ
Chemical Structures, in Silico Docking, and Sequence Analyses
Chemical structures and electrostatic surface potentials (range, −0.1 to 0.1) were plotted using Jmol, an open-source Java viewer for chemical structures in 3D: jmol.org. For docking, the X. laevis KCNQ1 cryo-EM structure49 was first altered to incorporate KCNQ3/KCNQ5 residues known to be important for RTG and ML-213 binding, and their immediate neighbors, followed by energy minimization using the GROMOS 43B1 force field50, in DeepView51. Thus, X. laevis KCNQ1 amino acid sequence LITTLYIGF was converted to LITAWYIGF, the underlined W being W265 in human KCNQ3/KCNQ5 and the italicized residues being the immediate neighbors in KCNQ3/KCNQ5. In addition, X. laevis KCNQ1 sequence WWGVVTVTTIGYGD was converted to WWGLITLATIGYGD, the underlined L being Leu314 in human KCNQ3/KCNQ5 and the italicized residues being the immediate neighbors in KCNQ5 and/or KCNQ3. Surrounding non-mutated sequences are shown to illustrate the otherwise high sequence identity in these stretches. Unguided docking of mallotoxin and other compounds to predict binding sites was performed using SwissDock52 with CHARMM forcefields53.
PTZ Chemoconvulsant Assay
We compared anticonvulsant activities of test compounds in male C57BL/6 mice (Charles River) aged 2-3 months. Mice were housed and used according to the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health (NIH Publication, 8th edition, 2011). The study protocol was approved by the Institutional Animal Care and Use Committee of University of California, Irvine, which confirmed that all relevant ethical regulations were adhered to. Chemicals were sourced from Sigma (St. Louis, Mo., USA). A pentylene tetrazole (PTZ) chemoconvulsant assay was used-. Mice were injected intraperitoneally with IVA and/or MTX, concentrations as indicated, or vehicle control (PBS with 1% DMSO), and then 30 min later injected intraperitoneally with 80 mg kg−1 PTZ. Following the PTZ injection, mice were caged individually and an observer (GWA) blinded to the drug used recorded, over 20 min, clonic and tonic seizure incidence, and seizure-related mortality.
Statistical Analysis
All values are expressed as mean±SEM. Chi-squared analysis was used to compare seizure-related mortality in mice. One-way ANOVA was applied for all other tests; if multiple comparisons were performed, a post-hoc Tukey's HSD test was performed following ANOVA. All p-values were two-sided. Statistical significance was defined as p<0.05.
Supplemental materials are available with the online version of Manville et al., Nature Communications (2018) 9:3845; DOI: 10.1038/s41467-018-06339-2.
REFERENCES
- 1. White, H. S. Epilepsia 44 Suppl 7, 2-8 (2003).
- 2. Kamatenesi-Mugisha, M. & Oryem-Origa, H. Afr Health Sci 5, 40-49 (2005).
- 3. Spinella, M. Epilepsy & behavior: E&B 2, 524-532 (2001).
- 4. Kukuia, K. K., et al. Pharmacologia 3, 683-692 (2012).
- 5. Kukuia, K. K., et al. J Pharm Bioallied Sci 8, 253-261 (2016).
- 6. Gangwar, M., et al. Biomed Res Int 2014, 213973 (2014).
- 7. Abbott, G. W. New Journal of Science 2014, 26 (2014).
- 8. Wang, H. S., et al. Science 282, 1890-1893 (1998).
- 9. Selyanko, A. A., et al. The Journal of Neuroscience 19, 7742-7756 (1999).
- 10. Singh, N. A., et al. Nature genetics 18, 25-29 (1998).
- 11. Charlier, C., et al. Nature genetics 18, 53-55 (1998).
- 12. Biervert, C., et al. Science 279, 403-406 (1998).
- 13. Rostock, A., et al. Epilepsy research 23, 211-223 (1996).
- 14. Tober, C., et al. European journal of pharmacology 303, 163-169 (1996).
- 15. Rundfeldt, C. European journal of pharmacology 336, 243-249 (1997).
- 16. Main, M. J., et al. Molecular pharmacology 58, 253-262 (2000).
- 17. Tatulian, L., et al. The Journal of Neuroscience 21, 5535-5545 (2001).
- 18. Tatulian, L. & Brown, D. A. The Journal of physiology 549, 57-63 (2003).
- 19. Garin Shkolnik, T., et al. JAMA Dermatol 150, 984-989 (2014).
- 20. Mathias, S. V. & Abou-Khalil, B. W. Epilepsy Behav Case Rep 7, 61-63 (2017).
- 21. Schroder, R. L., et al. Neuropharmacology 40, 888-898 (2001).
- 22. Blackburn-Munro, G., et al. CNS drug reviews 11, 1-20 (2005).
- 23. Dupuis, D. S., et al. European journal of pharmacology 437, 129-137 (2002).
- 24. Korsgaard, M. P., et al. The Journal of pharmacology and experimental therapeutics 314, 282-292 (2005).
- 25. Wickenden, A. D., et al. Molecular pharmacology 58, 591-600 (2000).
- 26. Igwe, K. K., et al. International Journal of scientific research and management (IJSRM) 4, 4123-4129 (2016).
- 27. Oliver, B. E. P. Medicinal Plants in Nigeria, (Nigerian College of Arts, Science and Technology, Lagos, Nigeria, 1960).
- 28. Manville, R. W., et al. Nature communications 9, 1847 (2018).
- 29. Kim, R. Y., et al. Nature communications 6, 8116 (2015).
- 30. Matschke, V., et al. Cellular physiology and biochemistry: international journal of experimental cellular physiology, biochemistry, and pharmacology 40, 1549-1558 (2016).
- 31. Zaika, O., et al. Biophysical journal 95, 5121-5137 (2008).
- 32. Schenzer, A., et al. The Journal of Neuroscience 25, 5051-5060 (2005).
- 33. Lange, W., et al. Molecular pharmacology 75, 272-280 (2009).
- 34. Turner, W. A New Herball, parts II & III. (Steven Mierdman, London, UK, 1562, 1568).
- 35. Grieve, M. A Modem Herbal: The Medicinal, Culinary, Cosmetic and Economic Properties, Cultivation and Folk-lore of Herbs, Grasses, Fungi, Shrubs, & Trees with All Their Modern Scientific Uses, (1971).
- 36. Bruno, J. J. & Ellis, J. J. The Annals of pharmacotherapy 39, 643-648 (2005).
- 37. Zakharov, S. I., et al. The Journal of biological chemistry 280, 30882-30887 (2005).
- 38. Zeng, H., et al. The Journal of pharmacology and experimental therapeutics 319, 957-962 (2006).
- 39. Gschwendt, M., et al. Biochemical and biophysical research communications 199, 93-98 (1994).
- 40. Burkhill, H. M. The useful plants of west tropical Africa, (Royal Botanic Gardens, Kew, London, UK, 1994).
- 41. Hasan, M. M., et al. Arg. Biomed Res Int 2014, 539807 (2014).
- 42. Patel, R. K., et al. J Pharm Anal 2, 366-371 (2012).
- 43. Donath, F., et al. Pharmacopsychiatry 33, 47-53 (2000).
- 44. Dom, M. Forschende Komplementärmedizin und Klassische Naturheilkunde 7, 79-84 (2000).
- 45. Stevinson, C. & Ernst, E. Sleep medicine 1, 91-99 (2000).
- 46. Eadie, M. J. Epilepsia 45, 1338-1343 (2004).
- 47. Bialer, M., et al. Epilepsy research 43, 11-58 (2001).
- 48. Frost, G., et al. Nature communications 5, 3611 (2014).
- 49. Sun, J. & MacKinnon, R. Cell 169, 1042-1050 e1049 (2017).
- 50. van Gunsteren, W. F. Biomolecular simulation: the GROMOS96 manual and user guide, (Vdf Hochschulverlag ETHZ, 1996).
- 51. Johansson, M. U., et al. BMC bioinformatics 13, 173 (2012).
- 52. Grosdidier, A., et al. Nucleic acids research 39, W270-277 (2011).
- 53. Grosdidier, A., et al. Journal of computational chemistry 32, 2149-2159 (2011).
- 54. Abbott, G. W., et al. Science signaling 7, ra22 (2014).
Synthetic gabapentinoids, exemplified by gapapentin and pregabalin, are in extensive clinical use for indications including epilepsy, neuropathic pain, anxiety, and alcohol withdrawal. Their mechanisms of action are incompletely understood, but are thought to involve inhibition of α2δ subunit-containing voltage-gated calcium channels. This Example shows that gabapentin is a potent activator of the heteromeric KCNQ2/3 voltage-gated potassium channel, the primary molecular correlate of the neuronal M-current, and also homomeric KCNQ3 and KCNQ5 channels. In contrast, the structurally related gabapentinoid, pregabalin, does not activate KCNQ2/3, and at higher concentrations (≥10 μM) is inhibitory. Gabapentin activation of KCNQ2/3 (EC50=4.2 nM) or homomeric KCNQ3* (EC50=5.3 nM) channels requires KCNQ3-W265, a conserved tryptophan in KCNQ3 transmembrane segment 5. Homomeric KCNQ2 or KCNQ4 channels are insensitive to gabapentin, whereas KCNQ5 is highly sensitive (EC50=1.9 nM). Given the potent effects and the known anticonvulsant, antinociceptive, and anxiolytic effects of M-channel activation, these findings support an unexpected role for M-channel activation in the mechanism of action of gabapentin.
Gabapentin (Neurontin) and pregabalin (Lyrica) are synthetic antiepileptic and antinociceptive gabapentinoid compounds originally designed as analogues of the neurotransmitter γ-aminobutyric acid (GABA), and both are in widespread clinical use (Calandre et al., 2016). However, the mechanisms of action of gabapentinoids, exemplified by gabapentin and pregabalin, are incompletely understood. [3H]-gabapentin binding was first described in membrane fractions from rat brain homogenates, and the target protein identified as the α2-δ subunit of voltage-gated calcium (Cav) channels. The findings were later recapitulated using porcine brain tissue, heterologously expressed α2-δ, and also with pregabalin; binding was found to be exclusive to α2-δ and 2 isoforms (Brown and Gee, 1998; Field et al., 2006; Fuller-Bicer et al., 2009; Gee et al., 1996). Gabapentin and pregabalin are generally considered inactive against canonical GABAA and GABAB receptors, despite their structural similarity to GABA (Ben-Menachem, 2004; Jensen et al., 2002; Lanneau et al., 2001; Stringer and Lorenzo, 1999; Taylor, 1997), although some investigators contend that there are some subtype-specific effects on GABAB receptors (Bertrand et al., 2003a; Ng et al., 2001; Parker et al., 2004). Binding of gabapentin and pregabalin to α2-δ is suggested to act therapeutically via impairment of Cav channel activity, thus reducing neuronal calcium currents (Stefani et al., 1998; Stefani et al., 2001), although others observed no evidence for gabapentin-induced changes in neuronal Cav activity (Rock et al., 1993; Schumacher et al., 1998).
We recently made the unexpected discovery that GABA can activate voltage-gated potassium (Kv) channels composed of heteromeric assemblies of KCNQ2 (Kv7.2) and KCNQ3 (Kv7.3) pore-forming a subunits (Manville et al., 2018). KCNQ (Kv7) channels comprise tetramers of a subunits, each containing six transmembrane (S) segments, organized into the voltage-sensing domain (VSD, S1-4) and the pore module (S5-6) (
Because of the structural similarities between gabapentinoids and GABA, and the known influence of the M-current in many of the disease states responsive to gabapentinoids (epilepsy, pain, anxiety, alcohol withdrawal) (Blackburn-Munro et al., 2005; Kang et al., 2017; Mason et al., 2018), we hypothesized that gabapentinoids might modulate KCNQ2/3 channels. Here, using electrostatic surface mapping, in silico docking studies, cellular electrophysiology and site-directed mutagenesis, we examined whether the two gabapentinoids in widespread clinical use (gabapentin and pregabalin) can modulate KCNQ2/3 channel function.
Materials and MethodsChannel Subunit cRNA Preparation and Xenopus laevis Oocyte Injection
cRNA transcripts encoding human KCNQ2, KCNQ3, KCNQ4, KCNQ5 (Kv7.2-Kv7.5) were generated by in vitro transcription using the T7 polymerase mMessage mMachine kit (Thermo Fisher Scientific), after vector linearization, from cDNA sub-cloned into plasmids incorporating Xenopus laevis β-globin 5′ and 3′ UTRs flanking the coding region to enhance translation and cRNA stability. cRNA was quantified by spectrophotometry. Mutant KCNQ2 and KCNQ3 cDNAs were generated with site-directed mutagenesis using a QuikChange kit according to the manufacturer's protocol (Stratagene, San Diego, Calif.) and corresponding cRNAs prepared as above. Defolliculated stage V and VI Xenopus laevis oocytes (Ecocyte Bioscience, Austin, Tex.) were injected with KCNQ channel α subunit cRNAs (5-10 ng). The oocytes were incubated at 16° C. in Barth's saline solution (Ecocyte) containing penicillin and streptomycin, with daily washing, for 2-5 days prior to two-electrode voltage-clamp (TEVC) recording.
Two-Electrode Voltage Clamp (TEVC)
TEVC recording was performed at room temperature using a OC-725C amplifier (Warner Instruments, Hamden, Conn.) and pClamp8 software (Molecular Devices, Sunnyvale, Calif.) 2-5 days after cRNA injection as described in the section above. The oocytes were placed in a small-volume oocyte bath (Warner) and viewed with a dissection microscope. Unless otherwise stated, chemicals were sourced from Sigma. Bath solution was (in mM): 96 NaCl, 4 KCl, 1 MgCl2, 1 CaCl2, 10 HEPES (pH 7.6). Gabapentin and pregabalin were stored at −80° C. as 1 M stocks in molecular grade H2O and diluted to working concentrations on each experimental day. The drugs were introduced into the recording bath by gravity perfusion at a constant flow of 1 ml per minute for 3 minutes prior to recording. Pipettes were of 1-2 MΩ resistance when filled with 3 M KCl. Currents were recorded in response to pulses between −80 mV and +40 mV at 20 mV intervals, or a single pulse to +40 mV, from a holding potential of −80 mV, to yield current-voltage relationships, current magnitude, and for quantifying activation rate. Deactivation was recorded at −80 mV after a single pulse to +40 mV, from a holding potential of −80 mV. Electrophysiology data analysis was performed with Clampfit (Molecular Devices) and Graphpad Prism software (GraphPad, San Diego, Calif., USA); values are stated as mean±SD. Raw or normalized tail currents were plotted versus prepulse voltage and fitted with a single Boltzmann function:
where g is the normalized tail conductance, A1 is the initial value at −∞, A2 is the final value at +∞, V1/2 is the half-maximal voltage of activation and V: the slope factor. Activation and deactivation kinetics were fitted with single exponential functions.
Chemical Structures and Silico Docking
Chemical structures and electrostatic surface potentials were plotted and viewed using Jmol, an open-source Java viewer for chemical structures in 3D: jmol.org. For in silico ligand docking predictions, the Xenopuslaevis KCNQ1 cryoEM structure (Sun and MacKinnon, 2017) was first altered to incorporate KCNQ3/KCNQ5 residues known to be important for retigabine and ML-213 binding, and their immediate neighbors, followed by energy minimization as we previously described (Manville et al., 2018) using the GROMOS 43B1 force field (van Gunsteren, 1996) in DeepView (Johansson et al., 2012). Thus, Xenopus laevis KCNQ1 amino acid sequence LITTLYIGF was converted to LITAWYIGF, the underlined W being W265 in human KCNQ3 and the italicized residues being the immediate neighbors in KCNQ3/KCNQ5. In addition, Xenopus laevis KCNQ1 sequence WWGVVTVTTIGYGD was converted to WWGLITLATIGYGD, the underlined L being Leu314 in human KCNQ3 and the italicized residues being the immediate neighbors in KCNQ5 and/or KCNQ3. Surrounding non-mutated sequences are shown to illustrate the otherwise high sequence identity in these stretches. No other KCNQ1 residues were changed in the model. Unguided docking of gabapentin and pregabalin, to predict native binding sites, was performed using SwissDock with CHARMM forcefields (Grosdidier et al., 2011a; b).
Statistical Analysis
All values are expressed as mean±standard deviation (SD). One-way ANOVA was applied for all other tests; if multiple comparisons were performed, a post-hoc Tukey's HSD test was performed following ANOVA. All P-values were two-sided. Statistical significance was defined as P<0.05.
ResultsSynthetic anticonvulsants such as retigabine and ML-213 exhibit negative electrostatic surface potential near their carbonyl oxygen moieties, a chemical property thought to be important for activation of KCNQ2/3 channels (Kim et al., 2015). We previously found that GABA also possesses this chemical property, whereas the excitatory neurotransmitter, glutamate (which cannot open KCNQ2/3 channels) does not (Manville et al., 2018). Here, we found that gabapentin exhibits a similar negative electrostatic surface potential pattern to that of GABA, whereas pregabalin does not (
We next tested the predictions using the Xenopus laevis oocyte expression and two-electrode voltage-clamp (TEVC) electrophysiology. Gabapentin potently activated heteromeric KCNQ2/3 potassium channels, even at low nanomolar concentrations (
Gabapentin began to activate KCNQ2/3 immediately upon wash-in, with the current augmentation taking ˜2 minutes to plateau. Gabapentin effects washed out relatively slowly (<50% washout after 2 minutes) but the gabapentin-augmented current was rapidly inhibited by washing in the KCNQ channel inhibitor, XE991 (50 μM) (
We next examined the effects of gabapentin on homomeric channels formed by neuronal KCNQ isoforms. At 1 μM, gabapentin activated KCNQ3* (an expression-optimized KCNQ3-A315T mutant that ensures robust currents (Zaika et al., 2008)) and KCNQ5, especially at subthreshold potentials. In contrast, KCNQ2 and KCNQ4 were insensitive to 1 μM gabapentin (
Canonical GABAA and GABAB receptors are generally considered to be gabapentin-insensitive (Jensen et al., 2002; Taylor, 1997); in addition, previous studies have concluded that Xenopus laevis oocytes do not express endogenous GABAA or GABAB receptors (Guyon et al., 2013). Furthermore, the gabapentin-activated currents in KCNQ2/3 expressing oocytes were completely inhibited by the KCNQ-specific inhibitor, XE991 (
A Gabapentin Binding Site on KCNQ Channels
We recently discovered that KCNQ3 and KCNQ5 are directly activated by the inhibitory neurotransmitter GABA, which binds close to the highly conserved S5 tryptophan, KCNQ3-W265 (Manville et al., 2018). In the current study, we show that gabapentin likewise activates KCNQ3 and KCNQ5, whereas the related gabapentinoid, pregabalin, does not. Substitution of KCNQ3-W265 with a leucine prevents activation by GABA and gabapentin, and impairs GABA binding (Manville et al., 2018). KCNQ3-W265 (and its equivalent in KCNQ2, 4 and 5) is also very important for binding of retigabine and structurally related anticonvulsants (Schenzer et al., 2005). This is thought to be because small molecules with a strong negative electrostatic surface potential close to a carbonyl/carbamate oxygen can hydrogen-bond with the W265 (Kim et al., 2015). Indeed, here we found that pregabalin lacks this exposed negative surface potential and neither in silico docks, nor activates KCNQ3. Our in silico docking studies for gabapentin position it near to W265 and close to where retigabine (Kim et al., 2015) and GABA (Manville et al., 2018) are predicted to bind, but not necessarily overlapping—although no conclusions should be drawn from the small differences in poses, and resolution of the exact pose would require structural analysis and/or further mutagenesis to map the entire binding site. We conclude that the W265-based binding site evolved to accommodate GABA and other endogenous metabolites and analogs of GABA, leading to sensitivity to modem synthetic anticonvulsants including retigabine and gabapentin. Interestingly, KCNQ2-5 all bind GABA but only KCNQ3 and KCNQ5 are activated by GABA or gabapentin (Manville et al., 2018); retigabine activates all four (but not KCNQ1, which lacks the equivalent W) but KCNQ3 is the most sensitive (Tatulian et al., 2001).
Mechanisms of Therapeutic Action
Gabapentin and pregabalin are in wide clinical use to treat a variety of disorders of the nervous system, including neuropathic pain and epilepsy. There is considerable overlap between the clinical indications for each drug (Alles and Smith, 2018; Calandre et al., 2016; Sills, 2006). This, together with the contrasting ability of gabapentin and pregabalin to activate neuronal KCNQ isoforms found herein, suggests that KCNQ activation cannot be the dominant mechanism of action for the majority of the therapeutic effects of gabapentin. Gabapentinoid binding to the α2δ-1 subunit reportedly inhibits α2δ-1-containing Cav channels (Stefani et al., 1998; Stefani et al., 2001), (Stefani et al., 1998; Stefani et al., 2001) although others found that gabapentinoids have little effect on Cav channel activity or Cav channel-dependent neurotransmitter release at presynaptic nerve terminals (Brown and Randall, 2005; Hoppa et al., 2012; Rock et al., 1993; Schumacher et al., 1998). α2δ-1-NMDA receptor complexes were recently discovered in human and rodent spinal cord; gabapentin inhibited α2δ-1-dependent potentiation of NMDA receptor activity and associated pain hypersensitivity, presenting a plausible mechanism for antinociceptive effects of gabapentin (Chen et al., 2018).
Multiple Gabapentinoid Targets in Neurons—a Role for KCNQs?
In a study comparing pregabalin and gabapentin effects on cultured dorsal root ganglion (DRG) neurons from neonatal rats, pregabalin and gabapentin produced biphasic effects (acute inhibition, but longer-term augmentation) on endogenous K+ currents. The enhancing effect was attenuated by pertussis toxin or by intracellular application of a synthetic cAMP analogue, suggesting an indirect mechanism involving G protein activation (McClelland et al., 2004). Another group also found that effects of gabapentin on inward rectifier K+ channels and N-type Ca2+ channels were pertussis toxin-sensitive (Bertrand et al., 2003b). Pertussis toxin is commonly used to inhibit the downstream effects of GABAB receptor activation, as it inhibits some (but not all) of the G proteins involved in this process (Asano et al., 1985). Yet, others have shown that GABAB receptors are insensitive to gabapentinoids (Lanneau et al., 2001), and GABAB receptor inhibitors did not alter the pregabalin-induced inhibition of Cav currents in neonatal rat DRG neurons (Martin et al., 2002; McClelland et al., 2004). The most likely explanation for this apparent discrepancy is that gabapentinoids can activate pertussis-sensitive G-proteins, but independent of GABAB receptors (Martin et al., 2002).
With respect to the DRG neuron K+ channel inhibition by pregabalin, it was apamin-sensitive implying it involved small-conductance Ca2+-activated K+ channels (McClelland et al., 2004). The K+ current enhancement did not begin until 10 minutes after initiation of administration of pregabalin, was apamin-insensitive, and was faster when pregabalin was applied intracellularly, suggesting an intracellular signaling mechanism. The gating kinetics and voltage dependence of the DRG Kv current described in the gabapentinoid study do not necessarily suggest against it containing an M-current component. Interestingly, KCNQ2 (which is gabapentin-insensitive) expression precedes that of KCNQ3 (gabapentin-sensitive) during human brain development (Tinel et al., 1998), and the effects of KCNQ channel inhibition upon depolarization-induced GABA release and action potential propagation also alter dramatically from P0-P7 in rat (Okada et al., 2003). Thus, in some neurons, M-current might be insensitive to gabapentin early in development (e.g., the first week), unless KCNQ5 was appreciably expressed. Furthermore, in the study of gabapentinoid action on DRG neurons, K+ channel activity was quantified at +40 mV, a voltage at which the activating effects of gabapentin (and most Kv channel activators) are minimal. In addition, we find that pregabalin inhibits KCNQ2/3 channel activity at concentrations of 10 μM and above, suggesting that at the concentrations used in the prior study (250 μM) (McClelland et al., 2004) pregabalin would inhibit KCNQ2/3 channels and may have similar effects on other KCNQ isoforms that could be expressed in neonatal rat DRG neurons.
It is highly possible, given the somewhat pleiotropic actions of gabapentinoids, that the potent effects of gabapentin on KCNQ3 and KCNQ5 channels might be masked by other effects observed at higher doses, both experimentally and with respect to clinical mechanisms of action. Serum gabapentinoid concentrations may reach 100 μM in patients (although in the brain and spinal cord this concentration is likely to be lower) (Ben-Menachem et al., 1992; Ben-Menachem et al., 1995; Berry et al., 2003), several orders of magnitude higher than the EC50 for gabapentin activation of KCNQ2/3, KCNQ3 and KCNQ5 channels, but as noted above, within the range for pregabalin inhibition of KCNQ2/3.
Gabapentin has also been found to augment KATP currents in rat hippocampal and human neocortical slices (but not incidentally, in rat DRG neurons) (Freiman et al., 2001), and to inhibit the hyperpolarization-activated, cyclic nucleotide-gated channel, HCN4, albeit not at clinically relevant drug concentrations (Tae et al., 2017). Conversely, gabapentin augmented in hippocampal and inhibitory interneurons, cells that highly express HCN1 and HCN2 (Peng et al., 2011; Surges et al., 2003). Thus, indirect modes of action of gabapentin may occur in vivo, as reported for Kv currents in rat DRG neurons (McClelland et al., 2004).
Conclusions
Perhaps the two most important take-home points from this study are, first, that we have discovered a new chemical space for KCNQ2/3 activation by synthetic compounds. Future structure-activity relationship studies guided by what we now know regarding the difference between gabapentin versus pregabalin with respect to KCNQ opening, and our previous work identifying endogenous activators for KCNQ3 and KCNQ5, including GABA, GABOB and β-hydroxybutyrate (Manville et al., 2018), can start to inform synthesis of a new class of KCNQ activators for potential therapeutic use. Second, the high potency but relatively low efficacy of gabapentin compared to, e.g., retigabine, suggests the possibility that gabapentin could act as a partial agonist and disrupt therapeutic actions of retigabine and related anticonvulsants. Furthermore, it is possible that gabapentin competes with the binding of endogenous GABA and its metabolites to neuronal KCNQ channels but shares similar or lower efficacy to them with respect to KCNQ activation, possibly explaining why KCNQ activation may not be an important determinant of gabapentin's beneficial effects. Thus, further exploration of gabapentinoids and related compounds with respect to KCNQ activation might uncover superior compounds, which either avoid KCNQ activation and thus potentially disruptive partial agonism, or alternatively are more effective than gabapentin in activating neuronal KCNQs and thus clinically superior because of an additional, beneficial target site.
REFERENCES
- Alles S R A and Smith P A (2018) Pharmacological reviews 70:315-347.
- Asano T, Ui M and Ogasawara N (1985) The Journal of biological chemistry 260:12653-12658.
- Ben-Menachem E (2004) Epilepsia 45 Suppl 6:13-18.
- Ben-Menachem E. et al (1992) Epilepsy research 11:45-49.
- Ben-Menachem E. et al (1995) Epilepsy research 21:231-236.
- Berry D J, et al (2003) Seizure: the journal of the British Epilepsy Association 12:28-36.
- Bertrand S, et al (2003a) Hippocampus 13:525-528.
- Bertrand S, et al (2003b) Synapse 50:95-109.
- Blackburn-Munro G, et al (2005) CNS drug reviews 11:1-20.
- Brown D A and Adams P R (1980) Nature 283:673-676.
- Brown J P and Gee N S (1998) The Journal of biological chemistry 273:25458-25465.
- Brown J T and Randall A (2005) Synapse 55:262-269.
- Calandre E P, et al (2016) Expert Rev Neurother 16:1263-1277.
- Chen J, et al (2018) Cell reports 22:2307-2321.
- Field M J, et al (2006) Proc. of the Nat'l Acad. Sci. USA 103:17537-17542.
- Freiman T M, et al (2001) Naunyn-Schmiedeberg's arch. pharmacology 363:537-542.
- Fuller-Bicer G A, et al (2009) Amer. j. physiol. Heart & circ. physiol. 297:H117-124.
- Gee N S, et al (1996) The Journal of biological chemistry 271:5768-5776.
- Grosdidier A, et al (2011a) Journal of computational chemistry 32:2149-2159.
- Grosdidier A, et al (2011b) Nucleic acids research 39:W270-277.
- Guyon A, et al (2013) The Journal of Neuroscience 33:11643-11654.
- Hoppa M B, et al (2012) Nature 486:122-125.
- Jensen A A, et al (2002) Molecular pharmacology 61:1377-1384.
- Johansson M U, et al (2012) BMC bioinformatics 13:173.
- Kang S, et al (2017) Neuropsychopharmacology 42:1813-1824.
- Kim R Y, et al (2015) Nature communications 6:8116.
- Lanneau C, et al (2001) Neuropharmacology 41:965-975.
- Manville R W, et al (2018) Nature communications 9:1847.
- Marrion N V, et al (1989) British journal of pharmacology 98:557-573.
- Martin D J, et al (2002) Neuropharmacology 42:353-366.
- Mason B J, et al (2018) Expert opinion on investigational drugs 27:113-124.
- McClelland D, et al (2004) BMC pharmacology 4:14.
- Ng G Y, et al (2001) Molecular pharmacology 59:144-152.
- Okada M, et al (2003) Epilepsy research 53:81-94.
- Parker D A, et al (2004) European journal of pharmacology 495:137-143.
- Peng B W, et al (2011) Neuroscience letters 494:19-23.
- Rock D M, et a) (1993) Epilepsy research 16:89-98.
- Schenzer A, et al (2005) The Journal of Neuroscience 25:5051-5060.
- Schumacher T B, et al (1998) Epilepsia 39:355-363.
- Sills G J (2006) Current opinion in pharmacology 6:108-113.
- Stefani A, et at (1998) Neuropharmacology 37:83-91.
- Stefani A, et at (2001) Epilepsy research 43:239-248.
- Stringer J L and Lorenzo N (1999) Epilepsy research 33:169-176.
- Sun J and MacKinnon R (2017) Cell 169:1042-1050 e1049.
- Surges R, et at (2003) Epilepsia 44:150-156.
- Tae H S, et al (2017) Frontiers in pharmacology 8:554.
- Tatulian L, et al (2001) The Journal of Neuroscience 21:5535-5545.
- Taylor C P (1997) Revue neurologique 153 Suppl 1:S39-45.
- Tinel N, et al (1998) FEBS letters 438:171-176.
- van Gunsteren W F (1996) Biomolecular simulation: the GROMOS96 manual and user guide, Vdf Hochschulverlag ETHZ.
- Wang H S, et al (1998) Science 282:1890-1893.
- Zaika O, et al (2008) Biophysical journal 95:5121-5137.
Herbs have a long history of use as folk medicine anticonvulsants, yet the underlying mechanisms often remain unknown. Neuronal KCNQ potassium channel dysfunction can cause severe epileptic encephalopathies that are resistant to modem anticonvulsants. This Example shows that cilantro (Coriandrum sativum), a widely used culinary herb that also exhibits antiepileptic and other therapeutic activities, is a highly efficacious KCNQ channel activator. Screening of cilantro leaf metabolites revealed that one, the long-chain fatty aldehyde (E)-2-dodecenal, activates multiple KCNQs, including the predominant neuronal isoform, KCNQ2/KCNQ3 (EC50, 60 t 20 nM) and the predominant cardiac isoform, KCNQ1/KCNE1 (EC50, 260±100 nM). (E)-2-dodecenal also recapitulated the anticonvulsant action of cilantro, delaying pentylene tetrazole-induced seizures. In silico docking and mutagenesis studies identified the (E)-2-dodecenal binding site, juxtaposed between residues on the KCNQ S5 transmembrane segment and 54-5 linker. The results provide a molecular basis for the therapeutic actions of cilantro and indicate that this ubiquitous culinary herb is surprisingly influential upon clinically important KCNQ channels.
Documented use of botanical folk medicines stretches back as far as recorded human history itself (1). There is DNA evidence suggestive of consumption of plants for medicinal use by Homo neanderthalensis 48,000 years ago (2, 3), and archaeological evidence of non-food use by Homo erectus or similar species up to 800,000 years ago, of herbs used in the modem era as folk medicines (4). Evidence for the efficacy of such medicines ranges from anecdotal to clinical trials; similarly, mechanisms and active compounds have been elucidated for some botanical medicines while for others the molecular basis of action is unknown (5, 6).
In many cases, botanical medicines of current or historical use are also currently consumed, often on a large scale, as foodstuffs or food flavoring. One example is cilantro (Coriandrum sativum), known as coriander in the UK (
Voltage-gated potassium (Kv) channels within the KCNQ (Kv7) subfamily are sensitive to activation by a range of small molecules, including synthetic drugs, neurotransmitters and metabolites (9-13). Kv channels, including the five isoforms within the KCNQ (Kv7) subfamily, are composed of tetramers of pore-forming (a) subunits each comprising 6 transmembrane segments (S). S1-4 comprise the voltage sensing domain (VSD); S5-6 comprise the pore module (
The KCNQ family exhibits a wide range of tissue expression and functional attributes; KCNQ channels are therefore highly influential in many aspects of human physiology. Heteromeric KCNQ2/3 and also KCNQ3/5 channels generate the muscarinic-receptor-inhibited M-current, a subthreshold Kv current that regulates neuronal firing; homomeric KCNQ2, KCNQ3 and KCNQ5 channels may also contribute (15-19). Activation of KCNQ1, KCNQ4 and KCNQ5 channels expressed in vascular smooth muscle reduces vascular tone (20). KCNQ1 is also expressed in human heart, the inner ear and a variety of epithelia (21); KCNQ4 is expressed in auditory neurons and hair cells (22) and, like KCNQ1 (23), is essential for hearing.
Given the sensitivity of KCNQ channels to various small molecules and their diverse expression and functional roles, many of which potentially match the purported therapeutic effects of cilantro, here we screened KCNQ channels for sensitivity to cilantro. We discovered that cilantro activates various KCNQ isoforms, and identified a single cilantro metabolite (and its KCNQ channel binding site) that underlies the KCNQ-activating and anticonvulsant properties of cilantro.
Materials and MethodsPreparation of Plant Extracts
Certified organic fresh cilantro (Coriandrum sativum) was sourced from Mother's Market and Kitchen (Irvine, Calif., US), and homogenized fresh using a blender (SharkNinja, Needham, Mass., US). We then performed a methanolic extraction (80% methanol/20% water) on the cilantro homogenate for 48 hours at room temperature on a rocking platform with occasional inversion of the bottles to more fully resuspend the extract. Following this, the extract was filtered using Whatman filter paper #1 (Whatman, Maidstone, UK) and then the methanol was removed by evaporation in a fume hood for 48 hours at room temperature. The extract was then centrifuged for 10 minutes at 15° C., 4000 RCF to remove remaining particulate matter, followed by storage at −20° C. On the day of electrophysiological recording, the cilantro extract was thawed and then diluted 1:100 in bath solution (see below) immediately before use.
Channel Subunit cRNA Preparation and Xenopus laevis Oocyte Injection
As previously described (12), we generated cRNA transcripts encoding human KCNA1, KCNE1, KCNE2, KCNE3, KCNQ1, KCNQ2, KCNQ3, KCNQ4 or KCNQ5 by in vitro transcription using the T7 polymerase mMessage mMachine kit (Thermo Fisher Scientific), after vector linearization, from cDNA sub-cloned into plasmids incorporating Xenopus laevis β-globin 5′ and 3′ UTRs flanking the coding region to enhance translation and cRNA stability. We quantified cRNA by spectrophotometry. We generated mutant KCNQ2 and KCNQ3 cDNAs by site-directed mutagenesis with a QuikChange kit (Stratagene, San Diego, Calif.) and prepared the cRNAs as above. We injected defolliculated stage V and VI Xenopus laevis oocytes (Ecocyte Bioscience, Austin, Tex. and Xenoocyte, Dexter, Mich., US) with KCNE and/or KCNQ cRNAs (5-20 ng). We incubated the oocytes at 16° C. in Barth's saline solution (Ecocyte) containing penicillin and streptomycin, with daily washing, for 3-5 days prior to two-electrode voltage-clamp (TEVC) recording.
Two-Electrode Voltage Clamp (TEVC)
We performed TEVC at room temperature using an OC-725C amplifier (Warner Instruments, Hamden, Conn.) and pClamp10 software (Molecular Devices, Sunnyvale, Calif.) 2-5 days after cRNA injection as described in the section above. For recording, we placed the oocytes in a small-volume oocyte bath (Warner) and viewed them with a dissection microscope. We sourced chemicals from Sigma unless otherwise stated. We studied effects of 1% cilantro extract and of nine compounds previously identified in cilantro extract. Vanillic acid and 3,4-Dihydroxybenzoic acid were each solubilized in 100% ethanol at a stock concentration of 100 mM; the other compounds were solubilized directly in bath solution. We initially screened for KCNQ2/3 channel activity using 100 μM concentrations of each of the nine components, then conducted dose responses where appropriate. Bath solution was (in mM): 96 NaCl, 4 KCl, 1 MgCl2, 1 CaCl2, 10 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (pH 7.6). We introduced 1% cilantro extract, or each of the nine cilantro components, into the oocyte recording bath by gravity perfusion at a constant flow of 1 ml per minute for 3 minutes prior to recording. Pipettes were of 1-2 MΩ resistance when filled with 3 M KCl. We recorded currents in response to voltage pulses between −120 or −80 mV and +40 mV at 20 mV intervals from a holding potential of −80 mV, to yield current-voltage relationships, current magnitude, and for quantifying activation rate. We analyzed data using Clampfit (Molecular Devices) and Graphpad Prism software (GraphPad, San Diego, Calif., USA); values are stated as mean±SEM. We plotted raw or normalized tail currents versus prepulse voltage and fitted with a single Boltzmann function:
where g is the normalized tail conductance, A1 is the initial value at −∞, A2 is the final value at +∞, V1/2 is the half-maximal voltage of activation and Vs the slope factor. We fitted activation and deactivation kinetics with single exponential functions.
Chemical Structures and Silico Docking
We plotted and viewed chemical structures and electrostatic surface potential using Jmol, an open-source Java viewer for chemical structures in 3D: jmol.org. For in silico ligand docking predictions of binding to KCNQ2, we first altered the Xenopus laevis KCNQ1 cryo-electron microscopy-derived structure (PDB 5VMS) (24) to incorporate KCNQ2 residues important for retigabine binding, and their immediate neighbors, followed by energy minimization as we previously described (12) using the GROMOS 43B1 force field (25) in DeepView (26). We then performed unguided docking of (E)-2-dodecanal to predict potential binding sites, using SwissDock with CHARMM forcefields (27, 28). We used a similar approach to simulate binding to KCNQ1 and KCNQ1/KCNE1, but instead employed closed and open state models of either channel previously developed by others (29).
PTZ Chemoconvulsant Assay
We quantified the anticonvulsant activities of (E)-2-dodecenal and tridecanal in male C57BL/6 mice (Charles River, Wilmington, Mass.) aged 2-3 months. The mice were housed and used according to the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health (NIH Publication, 8th edition, 2011). The study protocol was approved by the Institutional Animal Care and Use Committee of University of California, Irvine. The chemicals were sourced from Sigma (St. Louis, Mo., USA). We used a pentylene tetrazole (PTZ) chemoconvulsant assay as we previously described (30). We injected the mice intraperitoneally with (E)-2-dodecenal (2 or 20 mgkg−1) (with or without 2.5 mgkg−1 XE991) or tridecanal (20 mgkg−1) solubilized in PBS, or vehicle control (PBS), and then 30 minutes later we injected mice intraperitoneally with 80 mgkg-PTZ. Following the PTZ injection, the mice were caged individually and an observer (GWA) blinded to the experimental condition timed the latency to first seizure.
Statistical Analysis
All values are expressed as mean±SEM. One-way ANOVA was applied for all tests; all p values were two-sided. Where appropriate, we applied Tukey's correction for multiple comparisons.
ResultsCilantro Extract Activates Multiple KCNQ Isoforms
We performed a methanolic extraction (80% methanol/20% water) on fresh cilantro (Coriandrum sativum) (
Cilantro extract had no effect on water-injected control oocytes (
KCNQ2/3 heteromers are the primary KCNQ channel generating M-current in mammalian brain, and the main target of retigabine-class anticonvulsants (9, 13). Here, cilantro extract (1/100) was effective at negative-shifting KCNQ2/3 voltage dependence of activation (
The other KCNQ family member, KCNQ1, is retigabine-insensitive and is expressed in human heart and a variety of secretory epithelia including the gastric glands, colon and thyroid (21). KCNQ1 is notable for its functional diversity, largely endowed by formation of complexes with KCNE single-transmembrane spanning ancillary subunits. While homomeric KCNQ1 is functional, it is thought that in vivo KCNQ1 always complexes with KCNE subunits (21). In the human heart and inner ear, KCNQ1-KCNE1 complexes form the relatively positive- and slowly activating IKS current in ventricular myocytes (32, 33). In contrast, constitutively active KCNQ1-KCNE3 channels are expressed in the colonic epithelium basolateral membrane, where they regulate chloride ion secretion (34). KCNQ1 was also activated by cilantro extract (
Outside the KCNQ family, cilantro induced a relatively small (−5.9±0.9 mV) shift in the voltage dependence of activation of another neuronally expressed Kv channel, KCNA1 (Kv1.1) (
A Single Cilantro Metabolite Recapitulates Cilantro Activation of KCNQs
We next screened the predominant metabolites found in cilantro extract for KCNQ2/3 opening activity. Strikingly, out of 9 metabolites tested at 100 μM, only one activated KCNQ2/3—the 12-carbon fatty aldehyde, (E)-2-dodecenal. This specificity was remarkable given that closely related 10, 11 and 13-carbon aldehydes did nothing to KCNQ2/3 activity (
Examining effects on homomeric KCNQs, we found that (E)-2-dodecenal (100 μM) shares a similar efficacy profile to cilantro extract, i.e., it preferentially negative-shifted the ΔV0.5activation of KCNQ2 and KCNQ5 versus KCNQ1, KCNQ3 and KCNQ4 (
Cilantro confers a variety of beneficial effects (8), several of which could potentially involve KCNQ activation. Here, we focused on its effects as an anticonvulsant. Prior work showed that cilantro extract delays the onset (increases latency) of pentylene tetrazole (PTZ)-induced seizures in rats, without altering the overall incidence of clonic or tonic seizures (36). We compared the effects on PTZ-induced seizure latency in mice of (E)-2-dodecenal (2 and 20 mg/kg) versus the closely structurally related, KCNQ2/3-inactive, tridecanal (20 mg/kg). We found that, strikingly, (E)-2-dodecenal increased the latency to first seizure >threefold (p=0.0021; n=15-29) at 20 mgkg−1 and almost threefold at 2 mgkg−1 (p=0.06; n=16-20), while tridecanal (20 mgkg−1) had no effect (p=0.40; n=15), versus day- and age-matched controls for each cohort, injected with PBS (
Together with the effects we observed in vitro on KCNQ2/3 and other neuronal KCNQ isoforms, these data support the conclusion that (E)-2-dodecenal is a preeminent component of the anticonvulsant action of cilantro. Additionally, KCNQ3/KCNQ5 channels may also contribute to neuronal M-current and their dysfunction could participate in epilepsy and other hyperexcitability disorders (37, 38). Here, we found that KCNQ3/KCNQ5 channels are almost as (E)-2-dodecenal-sensitive as KCNQ5 (similar potency, ˜30% lower efficacy), in contrast to the insensitive homomeric KCNQ3* (Supplementary
(E)-2-Dodecenal Activates KCNQ2/3 Via a Binding Site Spanning S5 and the S4-5 Linker
(E)-2-dodecenal possesses negative electrostatic surface potential centered at its sole carbonyl oxygen (
To assess the validity of this prediction we tested the (E)-2-dodecenal sensitivity of KCNQ2/3 channels with mutation to leucine of KCNQ2-W236 and KCNQ3-W265 (
Studying the effects of the W and R mutants on homomeric KCNQ2 channels is problematic because of relatively low current magnitude, therefore instead we examined the effects of single mutants of KCNQ2 and KCNQ3 in the context of KCNQ2/3 complexes. Each of the S5 W mutants (KCNQ2-W236 and KCNQ3-W265) exerted a similar effect on the maximal shift in KCNQ2/3 ΔV0.5activation induced by (E)-2-dodecenal, reducing this to −9 mV, a value intermediate between that of wild-type and WL/WL double-mutant KCNQ2/3 channels (
In contrast, the KCNQ2-R213A mutation rendered KCNQ2/3 channels completely insensitive to (E)-2-dodecenal, while the equivalent KCNQ3-R242A mutant channel responded to (E)-2-dodecenal almost as much as wild-type KCNQ2/3 (similar potency, slightly lower efficacy) (
Overall, the W and R mutant data are consistent with the R being more influential than the W in terms of (E)-2-dodecenal binding/activation, and also with KCNQ2 channels being more sensitive than KCNQ3 to (E)-2-dodecenal. As homomeric KCNQ3* is insensitive to (E)-2-dodecenal (
These data support the docking predictions and suggest that (E)-2-dodecenal activates KCNQ2/3 channels by binding in a site spanning S5 and the S4-5 linker, specifically between (KCNQ2 numbering) W236 and R213, with KCNQ2-R213 being the most influential residue tested.
KCNE1 Impacts the KCNQ1 (E)-2-Dodecenal Binding Site
KCNQ1 lacks the KCNQ2-W236 equivalent, but possesses the KCNQ2-R213 equivalent (R243 in human KCNQ1); therefore we also investigated the mechanism of (E)-2-dodecenal binding in KCNQ1 channels, and how this might be impacted by co-assembly with KCNE regulatory subunits. While (E)-2-dodecenal had minimal effects on KCNQ1/KCNE2 (Supplementary
We first performed in silico docking to predict possible binding sites of (E)-2-dodecenal to KCNQ1 and KCNQ1/KCNE1 channels, using coordinates from previously published closed and open-state models (29) (
We tested the predicted influence of KCNE1 by comparing effects on (E)-2-dodecenal binding of an R243A mutation in KCNQ1 versus KCNQ1/KCNE1 channels using TEVC (
Our findings demonstrate for the first time a specific molecular basis for the anticonvulsant effects of cilantro, and show that this widely and frequently used food plant is highly effective at activating multiple isoforms within a clinically prominent family of human Kv channels. Aside from its noted anticonvulsant action, cilantro has historically been used to treat hypertension, digestive disorders, and has reported anti-inflammatory, antibacterial, analgesic and other potentially therapeutic properties (8). The tissue expression of cilantro-sensitive KCNQs suggests their activation could contribute to many of the reported therapeutic effects of cilantro.
The activity profile of (E)-2-dodecenal in delaying seizure onset is remarkably similar to that observed for cilantro extract in a previous study employing PTZ-induced seizures in rats (36). We observed anticonvulsant activity at 2 mgkg−1 (E)-2-dodecenal, equivalent to 11 μM, a concentration at which the (E)-2-dodecenal dose response curve for shifting the ΔV0.5activation of KCNQ2/3 saturates (
Furthermore, (E)-2-dodecenal, also known as eryngial, is the primary component of the essential oil produced from leaves of Eryngium foetidum, (culantro or Mexican coriander), a plant also used extensively as a food across Asia, Africa and the Caribbean, and utilized in folk medicine as an anticonvulsant and hypotensive, among other applications (43). Interestingly, the neuronally-expressed (44) channel KCNQ5 is also highly expressed in the vasculature (45, 46) and its activation reduces vascular tone, potentially reducing blood pressure (20). Here, we found cilantro extract and specifically (E)-2-dodecenal, previously identified as the predominant component of cilantro leaf oil (35), to be a highly efficacious KCNQ5 activator, suggesting a possible molecular basis contributing to the historical use of cilantro and culantro as folk hypotensives, an application that was recently verified using cilantro crude extract in animal studies (47). KCNQ5 gene variants (48) and aberrant splicing (49) that impair its function cause epilepsy (and severe intellectual disability), suggesting its activation could also contribute to the anticonvulsant action of cilantro (and (E)-2-dodecenal).
We also found that co-assembly with KCNE1 influenced the cilantro (and (E)-2-dodecenal)-sensitivity of KCNQ1. KCNQ1-KCNE1 complexes generate the IKs current that helps to repolarize ventricular cardiomyocytes. IKs is downregulated in heart failure, and this is suggested to contribute to increased risk of ventricular fibrillation in heart failure (50). It is fascinating, then, that cilantro extract was recently found to improve left ventricular function in heart failure (51). KCNQ1 also forms complexes with KCNE across along the gastrointestinal tract—with KCNE2 (in the stomach) and KCNE3 (in the colon, intestine and duodenum) (21). Here we found that KCNQ1-KCNE3 channel activity is potentiated by cilantro and by (E)-2-dodecenal. Stimulation of KCNQ1-KCNE3 activity by cilantro lower in the gastrointestinal tract would be expected to increase cAMP-stimulated Cl− secretion (34) and thus promote digestion and buffer the luminal environment to protect against damage from gastric acid effluent (52). Activation of KCNQ5 by cilantro could also contribute to its gut stimulatory properties, as KCNQ5 is also expressed in gastrointestinal smooth muscle (53) and its activation might therefore relax muscle, potentially being therapeutic in gastric motility disorders such as diabetic gastroparesis.
The KCNQ activation profile of (E)-2-dodecenal bears both similarities and differences to that of other KCNQ openers. We recently found that mallotoxin, from the shrub Mallotus oppositifolius, which is used in African folk medicine, also activates KCNQ1-5 homomers, prefers KCNQ2 over KCNQ3, and in docking simulations binds in a pose reminiscent to that predicted for (E)-2-dodecenal, between (KCNQ2 numbering) R213 and W236. As for (E)-2-dodecenal (
While retigabine does not activate KCNQ1, because it lacks the equivalent S5 tryptophan (39), (E)-2-dodecenal and mallotoxin both do. Although the KCNQ family-conserved arginine at the voltage sensor foot (R243 in KCNQ1) was essential for mallotoxin binding to KCNQ2/3, KCNQ1 and KCNQ1/KCNE1 channels (41), an arginine at position 243 was not essential for (E)-2-dodecenal activation of KCNQ1 or KCNQ1/KCNE1. However, docking simulations and shifts in potency and efficacy resulting from R243A mutation suggest (E)-2-dodecenal binds somewhat proximal to R243 and the S4-5 linker (
(E)-2-dodecenal is a trans-2,3-unsaturated fatty aldehyde that possesses a carbonyl oxygen at one end, which is predicted to provide a strong negative electrostatic surface potential in that region of the molecule (
In addition to the widespread use of cilantro in cooking and as an herbal medicine, (E)-2-dodecenal itself is in broad use as a food flavoring and to provide citrus notes to cosmetics, perfumes, soaps, detergents, shampoos and candles (59). Our mouse seizure studies suggest it readily accesses the brain and it is likely that its consumption as a food or herbal medicine (in cilantro) or as an added food flavoring would result in KCNQ-active levels in the human body; we found the 1% cilantro extract an efficacious KCNQ activator, and (E)-2-dodecenal itself showed greater than half-maximal opening effects on KCNQ2/3 at 100 nM (>10 mV shift at this concentration) (EC50, 60 t 20 nM). We anticipate that its activity on KCNQ channels contributes significantly to the broad therapeutic spectrum attributed to cilantro, which has seen it persist as a folk medicine for thousands of years throughout and perhaps predating human recorded history.
REFERENCES
- 1. Inskeep, R. R. (1969) South African Archaeological bulletin 24, 21-29
- 2. Hardy, K., et al. (2012) Naturwissenschaften 99, 617-626
- 3. Weyrich, L. S., et al. (2017) Nature 544, 357-361
- 4. Melamed, Y., et al. (2016) Proc. Nat'l Acad. Sci. USA 113, 14674-14679
- 5. Hoffmann, D. (2003) Medical Herbalism: the science and practice of herbal medicine, Healing Arts Press, Rochester, Vt.
- 6. Moerman, D. E. (2009) Native American Medicinal Plants, Timber Press, Portland, Oreg.
- 7. Cumo, C. (2013) Encyclopedia of cultivated plants: from Acacia to Zinnia, ABC-CLIO, Santa Barbara, Calif., US
- 8. Sahib, N. G., et al. (2013) Phytother Res 27, 1439-1456
- 9. Main, M. J., et al. (2000) Molecular pharmacology 58, 253-262
- 10. Manville, R. W., and Abbott, G. (2018) Molecular pharmacology 94, 1155-1163
- 11. Manville, R. W., and Abbott, G. W. (2018) Nature communications 9, 3845
- 12. Manville, R. W., P et al. (2018) Nature communications 9, 1847
- 13. Wickenden, A. D., et al. (2000) Molecular pharmacology 58, 591-600
- 14. Tatulian, L., et al. (2001) The Journal of Neuroscience 21, 5535-5545
- 15. Biervert, C., et al. (1998) Science 279, 403-406
- 16. Klinger, F., et al. (2011) NeuroImage 58, 761-769
- 17. Singh, N. A., et al. (1998) Nature genetics 18, 25-29
- 18. Tzingounis, A. V., et al. (2010) Proc. Nat'l Acad. Sci. USA 107, 10232-10237
- 19. Wang, H. S., et al. (1998) Science 282, 1890-1893
- 20. Yeung, S. Y., et al. (2007) British journal of pharmacology 151, 758-770
- 21. Abbott, G. W. (2014) New Journal of Science 2014, 26
- 22. Kubisch, C., et al. (1999) Cell 96, 437-446
- 23. Neyroud, N., et al. (1997) Nature genetics 15, 186-189
- 24. Sun, J., and MacKinnon, R. (2017) Cell 169, 1042-1050 e1049
- 25. van Gunsteren, W. F. (1996) Biomolecular simulation: the GROMOS96 manual and user guide, Vdf Hochschulverlag ETHZ
- 26. Johansson, M. U., et al. (2012) BMC bioinformatics 13, 173
- 27. Grosdidier, A., et al. (2011) Nucleic acids research 39, W270-277
- 28. Grosdidier, A., et al. (2011) Journal of computational chemistry 32, 2149-2159
- 29. Kang, C., et al. (2008) Biochemistry 47, 7999-8006
- 30. Abbott, G. W., et al. (2014) Science signaling 7, ra22
- 31. Etxeberria, A., et al. (2004) The Journal of Neuroscience 24, 9146-9152
- 32. Tyson, J., et al. (1997) Human molecular genetics 6, 2179-2185
- 33. Vetter, D. E., et al. (1996) Neuron 17, 1251-1264
- 34. Schroeder, B. C., et al. (2000) Nature 403, 196-199
- 35. Potter, T. L. (1996) Journal of Agricultural and Food Chemistry 44, 1824-1826
- 36. Karami, R., et al. (2015) Iran J Neurol 14, 59-66
- 37. Gilling, M., et al. (2013) Frontiers in genetics 4, 54
- 38. Lerche, C., et al. (2000) The Journal of biological chemistry 275, 22395-22400
- 39. Schenzer, A., et al. (2005) The Journal of Neuroscience 25, 5051-5060
- 40. Kim, R. Y., et al. (2015) Nature communications 6, 8116
- 41. De Silva, A. M., et al. (2018) Sci Adv 4, eaav0824
- 42. Panaghie, G., Abbott, G. W. (2007) The Journal of general physiology 129, 121-133
- 43. Paul, J. H., et al. (2011) Fitoterapia 82, 302-308
- 44. Schroeder, B. C., et al. (2000) The Journal of biological chemistry 275, 24089-24095
- 45. Mackie, A. R., et al. (2008) The Journal of pharmacology and experimental therapeutics 325, 475-483
- 46. Yeung, S., et al. (2008) British journal of pharmacology 155, 62-72
- 47. Jabeen, Q., et al. (2009) Journal of ethnopharmacology 122, 123-130
- 48. Lehman, A., et al. (2017) American journal of human genetics 101, 65-74
- 49. Rosti, G., et al. (2018) Eur J Med Genet doi.org/10.1016/j.ejmg.2018.10.007
- 50. Lau, E., et al. (2015) PloS one 10, e0122754
- 51. Dhyani, N., et al. (2018) J Diet Suppl, 1-14
- 52. Walker, N. M., et al. (2002) Gastroenterology 123, 531-541
- 53. Jepps, T. A., et al. (2009) Am. j. physiol. Gastrointestinal & liver physiology 297, G107-115
- 54. Miceli, F., et al. (2013) Proc. Nat'l Acad. Sci. USA 110, 4386-4391
- 55. Amato, G., et al. (2011) ACS medicinal chemistry letters 2, 481-484
- 56. Brueggemann, L. I., et al. (2014) Molecular pharmacology 66, 330-341
- 57. Wickenden, A. D., et al. (2008) Molecular pharmacology 73, 977-986
- 58. Kubo, I., et al. (2004) J Agric Food Chem 52, 3329-3332
- 59. Zviely, M. (2009) Perfumer & Flavorist 34, 26-28
Voltage-gated potassium (Kv) channel loss-of-function causes a variety of inherited disorders, but developing small molecules that activate Kv channels has proven challenging. As described above, the inhibitory neurotransmitter γ-aminobutyric acid (GABA) directly activates Kv channels KCNQ3 and KCNQ5. In this Example, finding that inhibitory neurotransmitter glycine is KCNQ-inactive, we re-engineered it in silico to introduce predicted KCNQ-opening properties, screened by in silico docking, then validated the hits functionally and mechanistically in vitro and in vivo. Attaching a fluorophenyl ring centered glycine surface negative electrostatic potential on its carbonyl oxygen, converting glycine to a low-nM affinity KCNQ channel activator that competes with GABA for KCNQ binding. Repositioning the phenyl ring fluorine and/or adding a methylsulfonyl group increased efficacy and switched the KCNQ activation isoform-selectivity. Combining KCNQ2- and KCNQ3-selective glycine derivatives synergistically potentiated KCNQ2/3 activation by exploiting its heteromeric composition. Thus, in silico optimization and docking, with real-world screening of only three compounds, facilitated glycine re-engineering into several novel, potent KCNQ isoform-selective activators.
Voltage-gated potassium (Kv) channel pore-forming a subunits are generated by a numerous and diverse gene family comprising 40 members in the human genome, separated into 12 subfamilies. Native Kv channels also contain regulatory subunits that shape their functional properties and further expand their diversity and functional repertoire. Kv channels are essential for a wide range of physiological processes, and in many cases little functional redundancy is observed even between seemingly closely related isoforms. Accordingly, disruption of specific Kv channel α or β subunits by inherited or sporadic human gene variants (or gene deletion in mice) is associated with a variety of disease syndromes, many of which are severe and often lethal1.
Despite in-depth knowledge of many of the physiological functions of specific Kv channels, and of the pathophysiological consequences of their disruption, therapeutic pharmacological targeting of the channels has been challenging. One of the reasons for this is that a great many Kv channel-linked disorders, or channelopathies, arise from loss of function. Direct correction of these requires, therefore, channel openers—a more difficult task than developing channel inhibitors or blockers.
One of the best known Kv channel openers is the drug retigabine (ezogabine). Retigabine activates neuronal KCNQ channels by negative-shifting their voltage dependence of activation2,3. Heteromeric KCNQ2/3 channels are particularly important in generating the muscarinic-inhibited M-current, a background Kv current that acts as a gatekeeper to limit aberrant neuronal firing4,5. Retigabine was the first Kv channel opener to reach the clinic, but was withdrawn in 2017 because of off-target side effects—it turns the sclera and skin blue6. However, it is effective at opening KCNQ2/3 channels and was clinically useful, mostly as an add-on therapy, in epilepsy. Since the development of retigabine, a new syndrome was recognized, termed KCNQ2 encephalopathy7. Caused primarily by sporadic, KCNQ2 loss-of-function mutations (as carriers tend not to reproduce), this disease is notable for severe developmental delays in addition to epilepsy. Clearly, new activators of KCNQ2 and many other Kv channels are needed, and this need will be acknowledged further as other Kv channelopathies are identified.
We recently made the surprising discovery that the predominant inhibitory neurotransmitter γ-amino-butyric acid (GABA) binds in a similar binding pocket to that of retigabine, and activates KCNQ3, KCNQ5, and KCNQ2/3 channels8. We also found that other metabolites, GABA analogs, and phytochemicals bind to a similar site, the majority also opening KCNQ channels by favoring their activation at more hyperpolarized membrane potentials8-10. KCNQ channels, and possibly other Kv channels, thereby possess a binding pocket that accommodates numerous types of small molecule activators. Glycine, which is structurally related to GABA, is also an inhibitory neurotransmitter. Here, after finding that glycine does not activate KCNQ channels, we re-engineered the glycine structure in silico to introduce known properties of KCNQ activators, and tested candidates using docking simulations. With minimal real-world functional screening, this led to discovery of a series of potent KCNQ channel openers, including a pair of activators that leverage isoform preferences to synergistically activate KCNQ2/3.
ResultsIn Silico Re-Engineering Glycine to Activate KCNQ Channels
Synthetic anticonvulsants such as retigabine possess negative electrostatic surface potential near their carbonyl oxygen moieties, a property found to be important for their activation of KCNQ2/3 channels11 (
In the simplest derivatives, surface negative potential was still skewed away from the carbonyl (
For the second in silico prediction phase, using SwissDock we performed unbiased docking prediction analysis of the glycine derivatives to a KCNQ1-KCNQ3 chimeric model8 based on the recent cryo-EM derived KCNQ1 structure12. We were especially interested in binding in the pocket lined on one side by the S5 tryptophan (W265 on KCNQ3) that is important for retigabine and GABA binding8,13, and on the other side by the S4-S5 linker-proximal arginine at the foot of S4 (R242 in KCNQ3) that is required for binding of phytochemicals such as mallotoxin to KCNQ channels14 and mutation of which in KCNQ2 causes benign familial neonatal convulsions15 (
4FPG Isoform-Selectively Activates KCNQ Channels
We next validated the in silico predictions using two-electrode voltage-clamp of homomeric neuronal KCNQ2-5 channel isoforms expressed in Xenopus laevis oocytes. By quantifying the hyperpolarizing shift in voltage dependence of KCNQ channel activation (ΔV0.5act) versus [4FPG], we discovered that, as predicted, 4FPG is a KCNQ channel opener. 4FPG most potently activated (i.e., negative-shifted the voltage dependence of activation of) KCNQ4 (EC50=49±12 nM), followed by KCNQ2 (EC50=69±31 nM) and KCNQ1 (EC50=90±20 nM), and had no effects on KCNQ3* or KCNQ5 (
Subtle Modifications to 4FPG Create Derivatives with Altered KCNQ Isoform Selectivity
KCNQ4 activation is an unwanted property of KCNQ2 activators because while KCNQ2 activation may be beneficial in KCNQ2-dependent and other types of epilepsy, KCNQ4 activation is thought to lead to urinary retention by opening KCNQ4 in bladder detrusor muscle18. In addition, activation of KCNQ1, a cardiac and epithelial KCNQ channel α subunit, might also cause unwanted off-target effects. We therefore further re-engineered 4FPG in silico, creating several single-halide glycine derivatives, all of which exhibited negative electrostatic surface potential close to a carbonyl oxygen (
The most closely related glycine derivative to 4FPG was 2-(fluorophenyl)glycine (2FPG), the only difference being the fluorine position on the phenyl ring. 4FPG and 2FPG are each represented in several conformations on the Zinc database (http://zinc.docking.org/) used for ligand selection, and so we compared docking of all conformations. All three 2FPG conformations docked primarily to the known binding pocket (
Strikingly, 2FPG was a potent KCNQ2 isoform-selective opener (EC50=322±139 nM) with negligible effects on KCNQ1, 3*, 4 or 5 (
3FMSG exhibited still different selectivity, activating KCNQ3* most potently (EC50=18±12 nM) followed by KCNQ5 (EC50=171±112 nM), with negligible effects on KCNQ1, KCNQ2 and KCNQ4 (
3FMSG Inhibits Glycine Receptor GLRA1
We next tested whether the glycine derivatives 2FPG and 3FMSG modulated the canonical glycine receptor, GLRA1. Neither compound activated GLRA1 at 1 mM, in contrast to glycine (1 mM) (
Glycine Derivatives Differentially Activate KCNQ2/3 Channels
As KCNQ2/3 heteromers are the predominant neuronal KCNQ isoform, we tested their sensitivity to the glycine derivatives. All 3 derivatives activated KCNQ2/3 (
2FPG and 3FMSG Occupy the KCNQ Neurotransmitter Binding Pocket
Docking poses predicted that 2FPG binds between the S5 tryptophan (W236 in KCNQ2) and the arginine at the foot of S4 (R213 in KCNQ2) (
3FMSG also docked in silico between the S5 tryptophan (W265 in KCNQ3) and the arginine at the foot of S4 (R242 in KCNQ3) (
2FPG and 3FMSG KCNQ Isoform Selectivity Arises Primarily from Functional Selectivity
To examine the mechanism of KCNQ isoform selectivity among glycine derivatives, we first assessed the combined effects of 2FPG and 3FMSG on homomeric KCNQ2 and KCNQ3 channels. A tenfold excess of 3FMSG (100 μM) subtly reduced the efficacy of 2FPG (10 μM) with respect to KCNQ2 activation by shifting the voltage dependence of 2FPG action such that efficacy at −40 mV was greatly reduced (although activation was still sufficient to hyperpolarize the oocyte membrane potential because effects were greater at more negative membrane potentials). In contrast, a tenfold excess of 2FPG (100 μM) did not alter the effects of 3FMSG (10 μM) on KCNQ3* activation at any membrane potential (
We next used a radioligand binding assay to quantify tritiated GABA binding to homomeric KCNQ2 and KCNQ3* channels expressed in oocytes. As we previously found8, GABA bound to both KCNQ2 and KCNQ3 (
As each compound can bind to both KCNQ isoforms, isoform selectivity must arise predominantly from functional selectivity, not binding selectivity. There are clear binding preferences, otherwise a tenfold excess of the non-activating compound would greatly reduce the efficacy of the activating compound, and that did not occur (
2FPG and 3FMSG Synergistically Activate KCNQ2/3 Channels
As 2FPG and 3FMSG preferentially activate different KCNQ2/3 channel subunits (KCNQ2 and KCNQ3, respectively) we tested their ability to synergistically activate KCNQ2/3 by leveraging their isoform preferences. We used equal concentrations of each compound to avoid possible competition for the same-isoform binding site (see
Investigating the synergy further, we dropped the concentrations to 1 μM each of 2FPG and 3FMSG and still observed synergy for KCNQ2/3 activation, increasing to a 40-fold increase in current at −60 mV (
We previously found that gabapentin is a potent activator of KCNQ3 and KCNQ5 but not KCNQ2 channels, and that it also activates KCNQ2/3 heteromers9. Accordingly, we also found here that gabapentin synergizes with 2FPG with respect to KCNQ2/3 activation (
Together with the data in
The canonical signaling action of the inhibitory neurotransmitter, glycine, is binding to two types of ligand-gated chloride channel—the strychnine-sensitive glycine receptor and the N-methyl-D-aspartic acid (NMDA) receptor. In the glycine receptor, glycine binding induces a conformational change that activates the channel; in NMDA receptors, glycine binding potentiates the effect of glutamate binding to the same receptor20,21. KCNQ potassium channels are primarily gated by changes in membrane potential via their voltage sensor. Yet, they are also highly sensitive to a variety of small molecules, many of which can favor channel opening at a given membrane potential. We recently discovered that the primary inhibitory neurotransmitter in metazoan nervous systems, GABA, binds to KCNQ3-W265 and the equivalent residue KCNQ5-W270 to activate these channels, and also KCNQ2/3 complexes. This activity was also observed for native M-current in PC12 cells and mouse dorsal root ganglion neurons. In contrast, the inhibitory neurotransmitter glycine (
While glycine and glutamate are structurally related to GABA, unlike GABA they do not exhibit negative electrostatic surface potential centered on the carbonyl glycine, an established property of KCNQ channel openers that activate via KCNQ3-W26511. Here, we used mapping of electrostatic surface potential and docking to in silico-engineer a glycine derivative with predicted KCNQ-opening properties, with the initial hit (4FPG) arising from addition of a fluorophenyl group to the glycine amide group. Interestingly, 4FPG also activated KCNQ1, which lacks the S5 tryptophan required for activation by, e.g., retigabine and GABA, suggesting 4FPG can also activate via the S4/5-proximal arginine also important for KCNQ2/3 activation (although we did not pursue KCNQ1 mutagenesis studies herein). Remarkably, even subtle changes such as moving the fluorine atom two spaces along in the ring completely altered the KCNQ isoform selectivity of the glycine derivatives. While we do not yet understand the channel structural determinants underlying this selectivity switch, the finding suggests a novel avenue in which to explore future druggable derivatives that lack KCNQ4 activity (a side effect ideally avoided for anticonvulsants because of inhibition of bladder emptying22), as we observed for 2FPG and 3FMSG (
We previously discovered that the heteromeric composition of KCNQ2/3 channels can be leveraged to potentiate the opening action of small molecules by combining two or more compounds with different KCNQ isoform preferences. In the prior study, this involved two components of a traditional anticonvulsant (mallotoxin and isovaleric acid, from the African shrub Mallotus oppositifolius) that preferentially activate KCNQ2, together with retigabine, which prefers KCNQ310. Here, we found that the principle holds for the glycine-based KCNQ activators, and also for the combination of KCNQ2-preferring 2FPG and gabapentin, a widely used analgesic that also exhibits anticonvulsant activity and which we previously found to isoform-selectively activate KCNQ3 and KCNQ59. The KCNQ2/3 synergy approach may hold promise as a strategy for avoiding the individual toxicities of some compounds by combining them at lower (potentially safe) concentrations with compounds with alternate KCNQ isoform preferences, also at lower concentrations.
In the present study, we were able to identify three novel and potent KCNQ activators derived from glycine entirely by straightforward in silico approaches using publicly available programs (Jmol and SwissDock) without any custom modifications. Subsequent electrophysiological and mutagenesis work validated both this specific in silico screening strategy and also gave some support to the docking predictions in terms of binding position, as the S5 tryptophan and S4-5 arginine residues lining the small molecule binding site were influential in determining effects in vitro. With respect to the deeply (within the membrane) positioned S4-5 arginine, the compounds docked close to the arginine in silico and then when we mutated the arginine in the isoform that is sensitive to 2FPG (KCNQ2) versus 3FMSG (KCNQ3), we diminished or lost sensitivity specifically to the respective drug. This suggests that the arginine residue either forms part of the binding site or impacts the way in which binding is translated into channel activation. This could possibly be because the drug binding disrupts interaction between the arginine and the cell membrane, or because mutating the arginine disrupts its interaction with the cell membrane. However, the arginine mutants do not greatly alter the voltage dependence of activation at baseline, suggesting against their mutation dramatically altering gating or voltage sensing per se at baseline, at least in KCNQ2/3 channels.
In addition, the results of 2FPG and 3FMSG wash-in and washout studies are consistent with these molecules entering a deep binding pocket from the external face, and not having to first cross the cell membrane and then access the binding site from the inner face of the cell membrane. The simplest conclusion is that the compounds bind somewhere proximal to both the S5 tryptophan and the S4-5 arginine. There is one additional caveat, however. KCNQ2/3α subunits are expected to exhibit domain-swapping, whereby the VSD of one subunit aligns with the pore module of the adjoining subunit. This may potentially result in mixed-isoform binding sites, which could complicate interpretation of results. Based on the data herein, we feel secure in stating that 2FPG and 3FMSG can each bind to both homomeric KCNQ2 and KCNQ3* channels, and that their isoform selectivity arises predominantly from the selectivity of their functional effects, and to a lesser extent their binding selectivity (the latter occurs, but cannot explain the lack of effects of 2FPG and 3FMSG on their non-preferred isoform even at high concentrations). Further, our data conclusively demonstrate that 2FPG and 3FMSG (or gabapentin) synergistically activate KCNQ2/3 channels by leveraging their isoform selectivity (both binding preference and functional selectivity) and the heteromeric composition of these channels.
The screening approach we used will be applicable to many other classes of small molecules with respect to predicting KCNQ channel activation, i.e., identify those compounds with the preferred chemical properties, dock to filter out predicted non-binders and then validate in vitro. The docking program was able to correctly predict lack of glycine activity and also predicted binding of 2FPG, 4FPG and 3FMSG, but at this stage the model and/or docking program we use are not sophisticated enough to predict KCNQ isoform selectivity. While it is relatively trivial once predicted KCNQ activity is identified to test each of the KCNQ homomers for sensitivity in vitro, an accurate system in which KCNQ4 activity, for example, could be used as a filter before in vitro screening, would be beneficial. With relatively few in silico screening steps and sufficient computing power, it may therefore be possible in the future to identify safe, potent KCNQ openers that lack KCNQ4 opening activity from massive commercially available chemical libraries.
Materials and MethodsChannel Subunit cRNA Preparation and Xenopus laevis Oocyte Injection
We generated cRNA transcripts encoding human KCNQ1, KCNQ2, KCNQ3, KCNQ4, KCNQ5 or GLRA1 (NM_001146040) (GenScript, Piscataway, N.J., USA) by in vitro transcription using the T7 polymerase mMessage mMachine kit (Thermo Fisher Scientific), after vector linearization, from cDNA sub-cloned into plasmids incorporating Xenopus laevis β-globin 5′ and 3′ UTRs flanking the coding region to enhance translation and cRNA stability. We quantified cRNA by spectrophotometry. We generated mutant KCNQ2 and KCNQ3 cDNAs by site-directed mutagenesis using a QuikChange kit (Stratagene, San Diego, Calif.) and prepared the corresponding cRNAs as above. We injected defolliculated stage V and VI Xenopus laevis oocytes (Ecocyte Bioscience, Austin, Tex. and Xenoocyte, Dexter, Mich.) with KCNQ channel a subunit (5-20 ng) or GLRA1 (20 ng) cRNAs. We incubated the oocytes at 16° C. in Barth's saline solution (Ecocyte) containing penicillin and streptomycin, with daily washing, for 2-5 days prior to two-electrode voltage-clamp (TEVC) recording.
Two-Electrode Voltage Clamp (TEVC)
We performed TEVC at room temperature using an OC-725C amplifier (Warner Instruments, Hamden, Conn.) and pClamp10 software (Molecular Devices, Sunnyvale, Calif.) 3-5 days after cRNA injection as described in the section above. For recording, oocytes were placed in a small-volume oocyte bath (Warner) and viewed them with a dissection microscope. Chemicals were sourced from Sigma, Matrix Scientific and Santa Cruz. (2-fluorophenyl) glycine, N-(3-fluorophenyl)-N-(methylsulfonyl) glycine and 2-(Trifluoromethyl)-DL-phenylglycine were each solubilized in bath solution at a stock concentration of 10 mM; 2-amino-2-(4fluorophenyl) acetic acid and 4-(trifluoromethyl)-L-phenylglycine were solubilized in 1M hydrochloric acid at a stock concentration of 10 mM. All stock solutions were diluted in bath solution on the day of experiments. KCNQ2/3 channel activation was screened for using either 30 μM or 100 μM concentrations of each of the six compounds, then dose responses were conducted as appropriate. Bath solution was (in mM): 96 NaCl, 4 KCl, 1 MgCl2, 1 CaCl2, 10 HEPES (pH 7.6). Compounds were introduced into the oocyte recording bath by gravity perfusion at a constant flow of 1 ml per minute for 3 minutes prior to recording. Pipettes were of 1-2 MΩ resistance when filled with 3 M KCl. Currents were recorded in response to voltage pulses between −120 or −80 mV and +40 mV at 20 mV intervals from a holding potential of −80 mV, to yield current-voltage relationships, current magnitude, and for quantifying activation rate. We analyzed data using Clampfit (Molecular Devices) and Graphpad Prism software (GraphPad, San Diego, Calif., USA); values are stated as mean±SEM. We plotted raw or normalized tail currents versus prepulse voltage and fitted with a single Boltzmann function:
where g is the normalized tail conductance, A1 is the initial value at −∞, A2 is the final value at +∞, V1/2 is the half-maximal voltage of activation and Vs the slope factor. Activation, deactivation, wash-in and washout kinetics were fitted with single exponential functions.
Relative Permeability Calculations
According to the Goldman-Hodgkin-Katz (GHK) voltage equation:
Where Erev is the absolute reversal potential and P is permeability. This permits calculation of the relative permeability of each ion if concentrations on either side of the membrane are known. A modified version of this equation was used here to determine relative permeability of two ions in a system in which only the extracellular ion concentration was known. Thus, relative permeability of Rb+, Cs+, and Na+ compared to K+ ions was calculated for KCNQ2 and KCNQ3* by plotting the IN relationships for each channel with each extracellular ion (100 mM) (using the voltage protocol shown in
These values were then compared for each channel against Rb+, Cs+, Na+ and K+ containing 100 μM 2FPG or 100 μM 3FMSG and statistical significance assessed using ANOVA.
GABA Competition Binding
Each group of oocytes was placed in a round-bottomed, 15-ml falcon tube, washed with ND96, and then resuspended in ND96 containing 10 μM γ-[2,3-3H(N)]-aminobutyric acid (3H-GABA) (Perkin Elmer, Waltham, Mass.) at 25-40 Ci/mMol specific activity either alone, or with 100 μM 2FPG or 100 μM 3FMSG, for a 30 minute incubation at room temperature. Oocytes were then washed four times in 16° C. ND96, transferred to individual wells in a 96 well plate and lysed in 0.2% SDS in ND96. Each oocyte lysate was transferred to a scintillation vial containing 5 ml Cytoscint scintillation cocktail fluid (MP Biomedicals, Santa Ana, Calif.). Vials were capped, shaken, and then allowed to sit at room temperature for at least 30 min before scintillation counting in a Beckmann Coulter LS6500 liquid scintillation counter.
Chemical Structures and Silico Docking
We plotted and viewed chemical structures and electrostatic surface potentials using Jmol, an open-source Java viewer for chemical structures in 3D: jmol.org. For in silico ligand docking predictions of binding to KCNQ2-5 channels, the Xenopus laevis KCNQ1 cryoEM structure (PDB 5VMS)12 was first altered to incorporate KCNQ3/KCNQ5 residues known to be important for retigabine and ML-213 binding, and their immediate neighbors, followed by energy minimization as we previously described8 using the GROMOS 43B1 force field23 in DeepView24. We performed unguided docking of the compounds described in the manuscript, to predict potential binding sites, using SwissDock with CHARMM forcefields25,26.
PTZ Chemoconvulsant Assay
We compared anticonvulsant activities of 3FMSG in male C57BL/6 mice (Charles River) aged 2-3 months. Mice were housed and used according to the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health (NIH Publication, 8th edition, 2011). The study protocol was approved by the Institutional Animal Care and Use Committee of University of California, Irvine, which confirmed that all relevant ethical regulations were adhered to. Chemicals were sourced from Sigma (St. Louis, Mo., USA). We used a pentylene tetrazole (PTZ) chemoconvulsant assay as we previously described27. Mice were injected intraperitoneally with 3FMSG (5 mg kg−1) or vehicle control (PBS) and then 30 minutes later injected intraperitoneally with 80 mg kg−1 PTZ. Following the PTZ injection, mice were caged individually and an observer (GWA) blinded to the drug used recorded seizure activity over 20 minutes. Tonic seizures are easily recognizable in mice as the hind legs stretch out behind the animal, following which the mouse remains prone for several minutes and often does not recover.
Statistical Analysis
All values are expressed as mean±SEM. One-way ANOVA was applied for the majority of tests; if multiple comparisons were performed, a post-hoc Tukey's HSD test was performed following ANOVA. For comparison of tonic seizure incidence, Fisher's exact test was used. All P-values were two-sided. Statistical significance was defined as P<0.05.
REFERENCES
- 1 Abbott, G. W. New Journal of Science, 26, doi: 10.1155/2014/237431 (2014).
- 2 Rundfeldt, C. European journal of pharmacology 336, 243-249 (1997).
- 3 Main, M. J. et al. Molecular pharmacology 58, 253-262 (2000).
- 4 Schroeder, B. C., et al. Nature 396, 687-690, doi:10.1038/25367 (1998).
- 5 Wang, H. S. et al. Science 282, 1890-1893 (1998).
- 6 Beacher, N. G., et al. BMC Oral Health 15, 122, doi:10.1186/s12903-015-0102-y (2015).
- 7 Weckhuysen, S. et al. Annals of neurology 71, 15-25, doi:10.1002/ana.22644 (2012).
- 8 Manville, R. W., et al. Nature communications 9, 1847, doi:10.1038/s41467-018-04266-w (2018).
- 9 Manville, R. W. & Abbott, G. Molecular pharmacology 94, 1155-1163, doi:10.1124/mol.118.112953 (2018).
- 10 Manville, R. W. & Abbott, G. W. Nature communications 9, 3845 (2018).
- 11 Kim, R. Y. et al. Nature communications 6, 8116, doi:10.1038/ncomms9116 (2015).
- 12 Sun, J. & MacKinnon, R. Cell 169, 1042-1050 e1049, doi:10.1016/j.cell.2017.05.019 (2017).
- 13 Schenzer, A. et al. The Journal of Neuroscience 25, 5051-5060, doi:10.1523/JNEUROSCI.0128-05.2005 (2005).
- 14 De Silva, A. M., et al. Sci Adv 4, eaav0824, doi:10.1126/sciadv.aav0824 (2018).
- 15 Castaldo, P. et al. The Journal of Neuroscience 22, RC199 (2002).
- 16 Etxeberria, A., et a. The Journal of Neuroscience 24, 9146-9152, doi:10.1523/JNEUROSCI.3194-04.2004 (2004).
- 17 Schroder, R. L. et al. Neuropharmacology 40, 888-898 (2001).
- 18 Svalo, J. et al. European journal of pharmacology 715, 312-320, doi:10.1016/j.ejphar.2013.05.005 (2013).
- 19 Manville, R. W., et al. Biophysical journal 113, 613-626, doi:10.1016fj.bpj.2017.06.055 (2017).
- 20 Johnson, J. W. & Ascher, P. Nature 325, 529-531, doi:10.1038/325529a0 (1987).
- 21 Valdes, F. & Orrego, F. Nature 226, 761-762 (1970).
- 22 Bientinesi, R. et al. Naunyn-Schmiedeberg's archives of pharmacology 390, 127-137, doi:10.1007/s00210-016-1312-9 (2017).
- 23 van Gunsteren, W. F. Biomolecular simulation: the GROMOS96 manual and user guide, (Vdf Hochschulverlag ETHZ, 1996).
- 24 Johansson, M. U., et al. BMC bioinformatics 13, 173, doi:10.1186/1471-2105-13-173 (2012).
- 25 Grosdidier, A., et al. Nucleic acids research 39, W270-277, doi:10.1093/nar/gkr366 (2011).
- Grosdidier, A., et al. Journal of computational chemistry 32, 2149-2159, doi:10.1002/jcc.21797 (2011).
- Abbott, G. W. et al. Science signaling 7, ra22, doi:10.1126/scisignal.2005025 (2014).
Throughout this application various publications are referenced. The disclosures of these publications in their entireties are hereby incorporated by reference into this application in order to describe more fully the state of the art to which this invention pertains.
Those skilled in the art will appreciate that the conceptions and specific embodiments disclosed in the foregoing description may be readily utilized as a basis for modifying or designing other embodiments for carrying out the same purposes of the present invention. Those skilled in the art will also appreciate that such equivalent embodiments do not depart from the spirit and scope of the invention as set forth in the appended claims.
Claims
1. A composition that activates KCNQ2/3, KCNQ3/5, or KCNQ4/5 voltage-gated potassium channels in a cell membrane, said channels comprising KCNQ2 subunits, KCNQ3 subunits, KCNQ4 subunits, and/or KCNQ5 subunits, the composition comprising:
- a first agent that binds a KCNQ2 subunit or a KCNQ5 subunit, and
- a second agent that binds a KCNQ3 subunit or a KCNQ4 subunit,
- wherein the first and second agents comprise non-identical agents.
2. The composition of claim 1, wherein the first agent that binds a KCNQ2 or KCNQ5 subunit is one or more agents selected from mallotoxin (MTX), isovaleric acid (IVA), zinc pyrithione (ZnPy), 2-fluorophenylglycine (2FPG), aloperine, 4-(fluorophenyl)glycine (4FPG), ICA-069673, ICA-27243, ztz-240, and (E)-2-dodecenal (E2D); retigabine (RTG) and derivatives thereof, gabapentin (GBP), β-hydroxybutyric acid (BHB), γ-Amino-β-hydroxybutyric acid (GABOB), and
- wherein the second agent that binds a KCNQ3 or KCNQ4 subunit is one or more agents selected from retigabine (RTG) and derivatives thereof, gabapentin (GBP), β-hydroxybutyric acid (BHB), γ-Amino-β-hydroxybutyric acid (GABOB), and N-(fluorophenyl)-N-(methylsulfonyl)glycine (3FMSG).
3. The composition of claim 1, wherein the first agent is MTX and/or IVA, and wherein the second agent is RTG.
4. The composition of claim 1, wherein the first agent is 2FPG, and wherein the second agent is GBP or 3FMSG.
5. The composition of claim 1, wherein the first and second agents are present in an amount ranging from 1 to 100 μM.
6. A method of activating KCNQ2/3, KCNQ3/5, or KCNQ4/5 voltage-gated potassium channels in a cell membrane, said channels comprising KCNQ2 or KCNQ5 subunits, and further comprising KCNQ3 or KCNQ4 subunits, the method comprising contacting the cell membrane with a composition of claim 1.
7. The method of claim 6, wherein the cell is a neuron, myocyte, epithelial cell, or endothelial cell.
8. The method of claim 6, wherein the cell membrane potential is −80 mV to +40 mV during the contacting of the cell membrane with the composition.
9. A method of reducing neuronal excitability in a subject, the method comprising administering to the subject a composition of claim 1.
10. A method of ameliorating symptoms of epilepsy, anxiety, neuropathic pain, hypertension, cardiovascular disease, a neurodegenerative disorder, alcohol withdrawal, cancer, inflammation, or ophthalmic disease in a subject, the method comprising administering to the subject a composition of claim 1.
11. The method of claim 9, wherein the first and second agents of the composition are each administered at a dose of 10-1000 mg/day.
12. The method of claim 6, wherein the channel is a KCNQ2/3 channel, and the combination of agents comprises one or more agents that bind KCNQ2 and one or more agents that bind KCNQ3.
13. The method of claim 12, wherein the agent that binds a KCNQ3 subunit is RTG, and wherein the RTG is administered at a dose of 20-240 mg/day.
14. The method of claim 12, wherein the agent that binds a KCNQ3 subunit is gabapentin.
15. The method of claim 6, wherein the channel is a KCNQ3/5 channel, and the combination of agents comprises one or more agents that bind KCNQ3 and one or more agents that bind KCNQ5.
16. The method of claim 15, wherein the agent that binds KCNQ3 is RTG, and the agent that binds KCNQ5 is aloperine.
17. The method of claim 6, wherein the channel is a KCNQ4/5 channel, and the combination of agents comprises one or more agents that bind KCNQ4 and one or more agents that bind KCNQ5.
Type: Application
Filed: Jul 18, 2019
Publication Date: Sep 2, 2021
Applicant: THE REGENTS OF THE UNIVERSITY OF CALIFORNIA (OAKLAND, CA)
Inventors: Geoffrey W. ABBOTT (IRVINE, CA), Rian W. MANVILLE (IRVINE, CA)
Application Number: 17/260,796
